Description
This is a report about Spectrum Analytics on Tracking Social Media ROI using.
Tracking Social Media ROI Viewing Spectrum Analytics
Marshall Sponder – Webmetricsguru.com
[email protected] @webmetricsguru 09/05/2010
pg 1
An Introduction to Spectrum Analytics for Social Media ROI
Social Media is a fairly new communications medium and activity that is often considered to be “free” and available to everyone. With abundant free and low cost tools and platforms to create, promote, monitor and analyze social media content businesses and individuals, excited about what Social Media can do for them, too often assume that measurement of Social Media can be taken care of as more of an afterthought. In the past estimating Social Media ROI was acceptable, increasingly that is no longer the case as outlined by TheBrandBuilder, Oliver Blanchard1. As CMOs and CEOs have started to seriously consider Social Media as an effective marketing medium, they also seeking a similar level of precision in Return on Investment (ROI) metrics they have gotten from Direct Marketing. Everyone is attempting to devise a framework for their company’s Social Media ROI, but the challenge is somewhat different for each business and industry. What are the metrics; beyond the basics such as how many Fans and Followers an account has that can actually help management manage better? And, of course, what metrics affect the HOLY GRAIL of business metrics: the bottom line? Lately, a lot of attention has been placed of formulas to calculate Social Media ROI. In fact, earlier this year Altimeter Group and Web Analytics Demystified shared a framework for Social Media Metrics and ROI2 addressing the Marketing Industry Challenge:”I can’t measure social media ROI.” The metrics were generalized to fit into a framework, which I analyzed3 in a series of posts at Webmetricsguru.com. I believe the framework is a step in the right direction, but much more guidance is needed. I will even go further by suggesting we avoid talking about
two dimensional formulas in order to calculate Social Media ROI, even though it is tempting to do so, before we are able to capture the information to put into those formulas. While we can reduce a business’s financials to a series of formulas that calculate profitability or loss, which is fairly well instrumented by accounting software and CPA’s and best practices have been for hundreds of years (for the sake of investment and profitability), it’s much more difficult to do the same thing for Social Media ROI (which is only a few years old).
Why I wrote this paper, finding Ultraviolet Data Nuggets and Rainbow Analytics
Over the past year I worked with Cecilia Pineda Feret4 analyzing the web site and social media of the Havana Central5 restaurant group with a few locations in Manhattan. We attempted to provide ROI information to the owner of the group and we regularly discussed the challenge of tracking our efforts for the company, the brand and the bottom line. For a small business, tracking and measuring this involves looking under the radar and following over the rainbow to find the ROI. Yet, every business has a different pot of golden data nuggets (or rainbow pancakes, they’re easier to find!), or RAINBOW ROI (Cecilia Pineda Feret coined this term in our discussions about Havana Central ROI tracking and analytics). The usual data on the radar are easy to capture: hits, followers, page views, etc. These metrics are more diagnostic according to Stratigent’s Jennifer Veesenmeyer6, allowing you to compare data points over time, and against competitors, and so on without necessarily providing any actionable insight into what the business should change or keep doing. Additional metrics are also required such as knowing results of efforts which were the most effective in bringing in sales, expanding the
pg 1
customer base, and increasing exposure of the brand. Acting as an analyst here my insights are usually connected with data I touch. I reasoned data to prove Social Media outreach is effective in driving new business and revenue was available, but we were not able to capture it. I termed such data as “ultraviolet”, because we can’t see ultraviolet light, yet it’s present all around us. I thought about what it takes to capture the information and tabulate it. When I started this line of thought my goal was not necessarily to solve Havana Central’s problem as much as to prove community management was an effective income generator for any restaurant chain. I decided to create an approach anyone could use to find how much data they were missing and what they could do to enable data capture in preparation for Social Media ROI based on the need to analyze the full spectrum of data and initial discussion of the concept can be found on my blog7. While coming up with this approach we realized Havana Central really needed a connection from the business and how it operates to the customer, and back to the business for tracking via analytics using tactics such as couponing, secret passwords and perks for VIPs, etc. Read on to find out more about the methodology. Similar diagnostics can be obtained from Google Analytics Benchmarking (when sharing data with Google) based on business type. The charts below show that on average, Havana Central gets 3.75 times more visits than the average restaurant using Google Analytics but the average restaurant is generating more pageviews per visit (it may have more to do, in this case, with PDF menus that are downloaded but not counted as pageviews or otherwise tracked by Google Analytics unless virtual pageviews are created for each menu). Havana Central’s Referral traffic as detailed by Compete. com shows no traffic at all from Social Networks (low sample site) and so a marketing goal ought to include meeting or surpassing the average referral traffic from Social Networks for the relevant category – in this case, several hundred more visits per month from Facebook alone would be called for (note: Google Analytics shows some Facebook and Twitter traffic to havanacentral.com). Figure 1 – Compete PRO Referral Analytics Traffic Dashboard, July 2010 for Industry Category: Entertainment Guide: Restaurants and Dining8
Setting Measurement Goals
Every business falls into some type of category. In order to know what is realistic to strive for in relation to Social Media, it helps to know what the average traffic is and sites for the category you business is in. For restaurants, on average 5% of referral traffic to top restaurants comes from Facebook, 0.16% from Twitter and .27% comes from MySpace.com according to Compete’s Referral Analytics Traffic Dashboard for the Entertainment Guide: Restaurant and Dining Category. (a new offering of Compete I provided suggestions about early on).
pg 2
for VentureBeat /SocialBeat last year10 suggested Video publishers had the IAB’s VAST standard to thank for rapid monetization of video ads that started taking place; I quote from my article here: … Monetizing online video has been somewhat problematic because the technology in place interfered Figure 2 – Google Analytics Benchmarking of Havanacentral.com vs. all other restaurant sites in Google’s database. But that’s beginning to change, partly due to the IAB Using benchmarking tools like Compete, Quantcast, Comscore, Nielsen, Alexa and in this respect, Google Analytics, is useful in order to formulate some business goals (which can also turn into measurement goals), especially those connected to Social Media, but they are not a replacement for direct measurement using site analytics. In fact, Web Analytics data and Compete.com data almost never match as the methods to obtain the information in both cases are different which accounts for the disparity in numbers of unique visitors, visits, pageviews, etc. In my discussions with Compete, they have also pointed out their tools are meant to complement, not replace site analytics. It’s also notable Comscore, Compete and Quantcast allow JavaScript tags to be added to your website which allow more accurate measurements for you business (but not necessarily your competitors unless they are also adding JavaScript tags to their sites). My work at the Web Analytics Association as Board Director in 2007-2009 led to creating and leading the Social Media Committee (now-defunct)11; it showed me the industry lacked consensus on social media standards and many of the players I talked to at the time liked it that The Web Analytics Association released Web Analytics Standards a few years back and there have been similar
9
with easy communications between publishers and third party video ad platforms.
(Internet Advertising Bureau), via the VAST Standard. According to the IAB, VAST (Video Ad Serving Template) includes a standard XML-based ad response for in-stream video and an XML Schema Definition (“XSD”) for developers. It’s meant to accommodate the majority of current practices within the online digital video advertising business. The gains we‘re seeing now were almost impossible to achieve until the VAST Standard, which made it easier to obtain end-to-end control between ad servers and associated video players (said CEO Tod Sacerdoti) Because we lack standard definitions for influence, conversations, engagement and conversions, to name a few, it’s so very much harder to reach an interoperable place where all the platforms that need to talk to each other, can, in a meaningful way.
Missing Measurement Standards
way (standards may be seen by some as a way to stifle innovation). I don’t know if that’s still true. On the other hand, for Social Media to “grow up” Standards definitions may be needed; I for one, believe they are very needed and important work is being done
pg 3
attempts by the Internet Advertising Bureau for Banner Advertising, Social Ads and Video Ads. Setting up standards may be a necessary step before widespread monetization can take place. In fact, an article I wrote
in the UK to achieve those standards through the CIPR (Certified Institute of Public Relations) Social Media Monitoring study group and panel , of which I’m an
12
person do Community Outreach, Management and track catering orders in a spreadsheet that could work for a very small operation that can live with a manual process but quickly becomes un-scalable. It also brings up another issue; at what point does tracking and monitoring roles interfere with going out and generating business? Marketing tasks are often assigned to different people/groups for reason; unless those people are the same page in every respect that matters how well can we expect them to communicate, or even agree on what needs to be done, needs to be tracked and how? The communications process is often too taxing for many
honorary member. I hope readers will follow CIPR-CP leader Philip Sheldrake quest to come up with standards
13
for Influence. In this paper the lack of standards could mean any measures a company comes up with to measure its own social media efforts (hopefully after aligning their business and measurement goals and strategies) will be difficult to apply universally, outside the context a specific business that came up with their own measures. That is OK for the short term, but in the longer term, we need consensus on what we measure and how to put that measurement in place. Once we have that, Rainbow Analytics will be much easier to achieve.
SMB’s and it’s difficult even for large corporations but larger businesses are usually willing to make strategic investments in messaging technology, business and social media measurement processes while SMB’s, often just trying to survive in difficult economic times, cannot or are unwilling to try. It might seem that communications problems (such as two individuals being on the same page) would be easy to solve, but it’s usually not the case. Another example from my experience working with Havana Central was creating and fine tuning Radian6 Alerts in order to know when customers were in a location and offering them a drink14 (or a coupon, which was never set up). I saw the potential here but got frustrated with the results. You can’t blame it on software, automation (Radian6, FourSquare and Facebook Places to name a few) took care of knowing people are in a venue by sending us email alerts (a measurement goal for Social Media attribution); but the business goal of Brand Awareness and Customer Loyalty (which are much harder to align and scorecard) was not aligned with the measurement strategy I came up with (getting management to be on the same page with business and measurement strategies is the subject of several books onto itself and I’m not going over it here).
pg 4
Aligning Business Processes to Measurement Goals
A common challenge for Social Media Measurement and the biggest roadblock finding the data nuggets under the radar and over the rainbow are business processes that are unaligned with a business’s goals for measurement. One example, obtaining Requests for a Quote (for a potential catering) which corresponds to a page on the Havana Central site (monitored by Google Analytics on one side and back end email request on the other) typically require communications between a marketing assistant and community manager (communications didn’t happen often enough or were too difficult to maintain). Catering jobs generate several thousand dollars of income, per job, to the bottom line, but the business process wasn’t yet aligned to the measurement goal (attribution of a referral to outreach). One solution is to have the same
rest of the organization). At least, that’s how it looked to Testing this approach in the real world of a NYC restaurant chain we found no systems in place yet to incentive very busy struggling restaurant managers, waiters and waitress (who need to be trained and often change jobs) to gather the information (a customer is in a location, greet them and offer a coupon) and record it in such a way that it can be attributed to Social Media and then tabulate into an ROI metric. Specific middleware to record the transactions was also missing. This lack of alignment is the biggest problem we have as the tools are in place to measure most of what we need already and plenty of people have come up with measures for Social Media ROI, but there’s almost succeeding in defining and aligning the social media dimension of your business goals to your overall measurement goals leading to the common result of Social Media ROI that cannot be accurately defined or measured (but is sometimes estimated); several people are working on this problem right now but no one I know of has solved it. Everyone reading this paper can likely attest to many instances of the very same alignment issues in their own organizations, once you step back and think about it. There may be several valid reasons why your business strategies aren’t aligned with measurement strategies; for example, when I worked at IBM.com (2003-2008) we were usually unable to track marketing and search optimization activities all the way down the sales channel. Part of the problem was traced back to technology (accounting systems that don’t speak to each other), another part were was sheer volume of activities and people (finding the right people in each part of IBM to talk to each other was problem) but the biggest issue did not turn out to be technology, as much as it was business processes for sales organization appeared to be different, and in conflict, with the measurement goals of the dotcom (IBM.com operates as a separate business unit from the
pg 5
me. Each part of an organization has its own goals and methodologies to serve its own purposes and to get work done; weather those purposes support or interfere with accurate measurement is up for you, the reader, to decide. In fact, just about every business unit, every company, every corporation, every city, state and every country have the same exact communication issues around goals and strategies, and as we have seen throughout our civilization’s history, this is not an easy problem to solve. Maybe the best way to look at our differences as stepping stone to better way of communication that serve us all (once there is enough consensus and people in place at the top that can focus business and measurement goals, strategies and policies).
Figure 3 – Radian6 charts showing Social Media Buzz Measurement of Havana Central Times Square location vs. neighboring restaurants Blue Fin, Heartland Brewery, Carmine’s Times Square, and Bond 45. The chart on the right shows the number of twitter followers of those who tweeted about each restaurant in the last 3 months and where they are located by region. Note: this chart shows aspects of Brand Awareness for the 5 restaurants surveyed here.
Why Social CRM is needed
Over the last year Social CRM has quickly evolved as means to connect Social Media Chatter to existing records in Salesforce and other CRMs. This space is quickly becoming one of the hot topics in Social Media Monitoring and will be subject of a future paper by the author.
Finding the right middleware to record results
This lack of alignment is the biggest problem we have as the tools are in place to measure most of what we need already and plenty of people have come up with measures for Social Media ROI, but there’s almost no one succeeding in defining and aligning the social media dimension of your business goals to your overall measurement goals leading to a common result of Social Media ROI that cannot be accurately defined or measured (but is sometimes estimated); several people are working on this problem right now but no one I know of has solved it.
Closing the Gap and Finding the Rainbow
Data can be collected from Social Media Monitoring platforms like Radian6 easily enough, but connecting a visitor (a social media awareness goal & a measurement goal) with increased sales by restaurant location generated from social media customers as a result of social media outreach (a business goal) takes a level of middleware (Social CRM, covered in the previous section) plus more buy in and patience from management (it’s often a hard combination to come by). Also needed is a way of connecting monitoring platforms such as Radian6 and the backend check receipts; today is only being envisioned and no one that I know of has done it yet. While this paper focuses on restaurants with its focus on location, check-ins, cash receipts, catering and reviews, the same process can be applied to any business with slight modifications. Barcode / QR code readers, RFID and Geo-fencing are expected to close the gap for restaurant venues in the next two years but the software to connect all this data must be
Figure 4 – part of the authors’ visualization of a Radian6 date export of visits by Twitter Accounts to any Havana Central location from June 09-June 10 (anyone checking in via FourSquare, or tweeting they were eating at one of the locations), seehttp://www.webmetricsguru.com/category/ social-media-case-study-havana-central/ for more details.
CRM aware and I expect the role of community manager will evolve into Relationship Management within the next 18 months. Acquisitions by Scout Labs by Lithium Technologies, Radian6 partnerships with Salesforce and recent enhancements by Alterian around the SM2 platform are just the beginning of the tip of the data iceberg lying
pg 6
under the surface, just waiting to be tapped.
example, each tweet about eating at the restaurant can be estimated to be worth over $100 based on the average order size, the average party size and the average price of a meal). Estimating Social Media ROI is useful when you make a
Getting Started by Performing a Social Enablement Audit
The only way to know how much precious data your business has in the unseen spectrum is to audit your data. Specifically, list all the sources of data your business has and what campaigns and marketing initiatives your business is running. This table shows data from Havana Central.
decision to expand or retract a marketing initiative but is not a replacement for actual ROI measures that are tied to a transaction. In another case, traffic and possible marketing efforts directed at the VIP page of the Havana Central restaurant site15 can be expected to appear in Google Analytics and Google Adwords (when ads are run) but not anywhere else.
Figure 6 – How enabled are we for Social Media ROI? Figure 5 – Social Enablement Audit of Havana Central I put all the data sources vertically and all the campaigns, including website promos and pages horizontally. Note, the x- and y-axis could be reversed easily. By conducting an audit of all the available data sources within in the context of all the known marketing campaigns and initiatives, we learn what information we have, where to find it, what format it is in and the level of difficulty in overlaying data in order to create Social Media ROI metrics we need. For Social Media ROI calculations that are tied to an actual In this example, data validating community management shows up partly in Google Analytics and Radian6, but will fail to appear anywhere else. It is next to impossible without a guaranteed tie into a revamped business process to track the actual dollar value to the outreach being done in behalf of the business, though one could estimate it (for transaction that takes place at the restaurant or its online site we need a logical way to overlay data, using what is commonly called a common key18, that currently doesn’t exist in this business context. Figure 6 suggests, based on the Enablement Audit that 82% of the data is “ultraviolet” or invisible (or, not able to be combined because a
pg 7
Most data is collected in Silos16, it’s pretty much always been that way and is getting worse according to Robert Scoble17. Applications are designed to provide specific functionality and data are collected around it, but usually applications aren’t designed to integrate with anything else – in this case, we don’t expect a reservation site such as Opentable, handling most restaurant reservations to have data on Community Management outreach of the VIP page of a website.
common key does not exist). One approach used by large corporations is to overlay data into Data Cubes or data warehouse where all the data you have can be imported into a Cube19, but usually the resources to pull something like this together are beyond reach of most businesses, including those who want measure ROI the most, small businesses and even PR Companies doing social media programs as part of Public Relations and Marketing (in behalf of their clients). Brick and mortar SMBs who could not possibly conceive, afford or execute such a data cube even though they may actually require one, but once your data is in the Cube you can perform analysis that could uncover ROI and customer behavior that is not ordinarily visible. Figure 8 – Data Coverage In the process of doing a Social Enablement Audit you might find that you are not collecting all the data you needed; in this case you may not have some data that you need in any form and in any data source. The best way to find out if you missing collecting data is to compare your data sources to each other. As you examine your data you may find there are some things you want to know that no application is tracking at all and this will drive new solutions for your business. What we’re trying to find out here is how compatible is the data being used to run your business to each other and how easy is it to integrate information. Typically, applications are not created to integrate and the data within each source and Figure 7 – slices of a data cube Given the lack of the needed common key (in this case, a common record locator for every record in Google Analytics, Opentable, SeemlessWeb, Email Marketing, etc) in many data sources – information needed to populate the data cube is going to be spotty, noisy and in other cases hard to translate it to the right structure to be effective. According to Gary Angel, the CTO of Semphonic, when you integrate data into a data warehouse, you open up new questions, targeting opportunities, and analysis methods that otherwise doesn’t exist19. are frequently incompatible with each other. One attempt to answer the compatibility problem was Adobe (formally Omniture) Open Business Analytics Platform20 using Web Analytics as central point to unify disparate applications such as Radian6 (one of our data sources), several different Content Management Systems, Salesforce Customer Relationship Management (CRM) and even Opentable, using Omniture for their own marketing analysis, (using Test and Target) within Site Catalyst, but all of these require a lot of money and time that do not make sense to implement for most small business (not until someone comes up with a suite of tools for low end of the market that makes data integrations much easier and inexpensive).
pg 8
Supposing you could integrate all the data you need into a cube, would the resulting reports lead to insights not ordinarily obtainable? I think so, but it will depend on how well business processes are translated in to the matching reports via a clearly defined measurement strategy, mentioned earlier, and gifted analysts needed to make sure they do. Hopefully, business will also learn how to listen to what the analysts have to say when they are able to come up with marketing insights and give them a seat at the table which they traditionally haven’t had in most organizations. I believe it’s time to consider doing similar enablement for Social Media that we more typically perform for Web Analytics (with extensive page tagging often required to pick up specific customer behavior, often a measurement goal).
could use the data to analyze their sites, but most businesses didn’t know how to define their business goals and translate them into measurement goals and align their organizations so that real analysis could happen with the result good analysis was done, but rarely. Around 2005 things got to be more interesting with a convergence of sorts including Flash and Ajax tracking evolving, but again, most organizations aren’t yet able, willing or think it’s too expensive to translate what they do and what they want to track into what these tools can do and what they can track, and as a result much data and insight is lost. As social media began as grassroots, peer to peer expression of communications between friends and fans, etc, this kind of integration had not lent itself to being instrumented in the same way Site Analytics evolved but it’s only a matter of time until it does get instrumented, much as Web Analytics has. As corporations are going more social, so are SMB’s. Parallels between Social Media Measurement and Web Analytics suggest convergence and the possibility of actual ROI measurement once business goals are merged with right measurement goals. With geo-location coming into the mix and checking in quickly becoming the universal currency, it may become easier to integrate social media with the rest of the data
Figure 9 – Stages of Web Analytics Evolution from Judah Phillips deck on Webanalytics congress the Netherlands, Amsterdam June 2nd 2010, slide 14 .
21
sources – and the cost and complexity will decrease on the low end, making possible for SMB’s to have a plug and play solution within the next two years. However, translating your business process, connecting the Data Rainbow back to measurement will be required even in streamlined solutions, leaving room for a lot of creativity.
In the beginning of Web Analytics circa early 1990’s basic diagnostic data is all that could be collected and today, many businesses are capable of tracking much more than this (with proper enablement) but settle for using the basic gauges of visitors, visits, pageviews though little insight into business can be obtained this way. At the turn of the century analytics improved and business
Finding Data Nuggets of Gold under Data Rainbow
It’s my belief we do not achieve actual Social Media ROI
pg 9
unless we can segment social media activity and associate it across data sources and marketing channels back to actual revenue. In fact, most business and consultancies have no idea how to start, which is why it doesn’t happen often, if it does at all.
pancakes!)
In order to know what can and can’t be combined to form Social Media ROI we need to look at the data we’re collecting in greater detail – assessing what we actually have to work with.
Figure 11 – A Opentable Report An Opentable report as it is currently generated is almost impossible to attribute to Social Media outreach there’s no common key to associate with a visitor from Facebook (unless the application tracks that). Opentable tracks website visitors that came from Havana Central, but there’s no way to count how much of it was Social Media related in the reports they provide to business owners. The same problem arises with Grubhub, a popular Restaurant Search utility.
Figure 10, Example of a Data Rainbow Only a fraction of the data we need to collect to track Social Media ROI is actually being saved (or seen) in a way that can be analyzed and connect to Social Media Marketing initiatives. The chart above shows some of the individual silos of data for Havana Central (the different data sources above might can be termed silos for the purpose of this paper). Figure 12 – A Grubhub report The Grubhub report has no common key except a date, everything else is tied back to internal databases that are not accessible (i.e.: ID number), no easy way to tie back to Google Analytics (unless we can put tracking code into a Grubhub instance, which is really hard to do for third party applications like Grubhub and Opentable).
Looking at the Data in our Silos for data nuggets (or rainbow
pg 10
Radian6 reports that monitor Buzz around the restaurant do offer possibilities to tracking what website visitors do on your website through its integration with WebTrends and in your business location via Salesforce.com but few businesses have adopted those integrations to my knowledge. As these Social Media targeted integrations exist at all but are rarely adopted, suggest businesses are too often unaware of the choices they have before them when they formulate their business goals and measurement strategies and often choose unwisely or fail to choose at all. In any case, while your Social Media campaigns might be driving a certain level activity which should be measureable, that same activity is invisible to the reports you can currently pull from Opentable and Grubhub, Google Analytics, Facebook and even Radian6 without deliberate measurement enablement and the alignment of your measurement strategy with you business strategy.
used to link all the transactions together it’s still possible to fill the gaps with third party applications and ingenuity.
Finding Solutions to Bridge the Gaps
Figure 14 – Measurement Solutions by Campaign or Marketing Initiative Taking a closer look at a specific marketing initiative against the sources of data available can drive solutions where missing data can be added in when it’s missing. Look at your campaigns and marketing. Figure out how to add the additional enablement that would allow you to track the Social Media ROI of a particular activity such as somebody downloading a menu if they came from Facebook or Twitter.
Comparing data dimensions and combining them in the Data Rainbow
Bridging the Data Gap Example – Connecting Facebook and Opentable for Data Nuggets
Figure 13 – Data Dimensions comparison Figure 15 – Finding ways to connect Opentable with In the chart above I highlighted in green dimensions from different data sources that might be able to be combined to get that Rainbow view and insights derived. We know by now the vast majority of data being collected cannot be combined based on the example above. While your data will too often be missing that common key that can be You may need to develop or find 3rd party applications capable of bridging the gaps found using this audit process. While we can enable Google Analytics/FB connection –we still need a Content Management System
pg 11
Facebook
that connects a website, Google Analytics and Opentable and Google Analytics can act as a go between.
2) Most businesses, esp. SMB’s are not yet ready to fully implement the data audit methodology. Someone needs to come along and make the process easier for them, and more streamlined, before this is going to happen on an SMB level. I suspect a streamlined implementation of data audits will differ somewhat based on business verticals. However, within each vertical there will be enough similarities to standardize and streamline (i.e.; most restaurants use Opentable and SeemlessWeb and should be part of the solution for this vertical, but perhaps not for other types of retail establishments like Supermarkets, clothing outlets, Travel Sites, Hotels, etc, which will need different sets of data feeds and metrics). 3) Depending on an actual SMBs’ goals, what you are measuring may be different for another SMB, even in the same or another industry. The challenge here is to capture the data that is not immediately seen or considered and
BTW, in the process of detailing needed capabilities our audit concludes the best enablement would be rebuilding that website (ouch!) with a CMS (SiteWrench CMS) that can handle Opentable & Google Analytics integration22.
relatable back to a Social Media effort by linking your measurement strategy and goals with your business strategy and goals. 4) To capture this data, we need to perform a complete audit of efforts made. It may be a time consuming task depending on the size and complexity of the business, but for the moment remains the best way to fully consider and fully integrate efforts across multiple platforms. 5) The best place to integrate all the data sources for
Conclusions:
1) “States of Readiness” for Social Media ROI Analytics implementation are very similar to those for Web Analytics, where there are various levels of participation using the tools currently available. For example, Radian6, Salesforce CRM and WebTrends integrate but most businesses are not ready to take advantage of it yet. Once business moves from past collecting basic diagnostic data, they can begin to merge their business strategy and goals with specific measurement strategies and goals and the measurement of Real Social Media ROI can begin.
your business is within the analytics platforms such as Google Analytics, Adobe Site Catalyst (formerly known as Omniture) though other solutions such as Spredfast are available. 6) You need to work with your web developers, programmers, product and sales teams to enable tracking of certain actions to ensure that your efforts are being tracked and what meaning you assign to them other than just collecting them on a periodic basis like many continue to do.
pg 12
4.http://www.linkedin.com/in/ceciliapinedaferet Cecilia Pineda Feret’s LinkedIn page. 5.http://www.havanacentral.com Havana Central Website 6.http://www.stratigent.com/news-and-events/pimp_ your_reports/default.html Stratigent’s Pimp Your Reports program which I recently attended 7.http://www.webmetricsguru.com/archives/2010/08/ on-360-degree-analytics/ - put forward that most of the interesting data that proves Social Media is working is invisible to the naked eye – or “ultraviolet”. However, with the right devices and mythology the data can be seen and quantified 8.http://www.compete.com/referrals/referrals/ category/3037/?threshold=1 What you’ll find at the end of the Data Rainbow …… Yum! (Amanda -http://iammommy.typepad.com/my_ weblog/2009/10/rainbow-pancake-sugar-cookies-winniethe-pooh.html) 10.http://social.venturebeat.com/2009/11/29/brightrollsays-video-ad-profits-are-soaring/ 11.http://www.webanalyticsassociation.org/?c_ 1.http://thebrandbuilder.wordpress.com/2009/05/29/howto-not-calculate-social-media-r-o-i/ Oliver Blanchard has written several posts and a popular SlideShare presenting his views on what is and is not Social Media ROI. 2. Altimeter Report: Social Marketing Analytics (Altimeter Group & Web Analytics Demystified) seehttp://www.webstrategist.com/blog/2...ing-analytics-with-web-analytics-demystified/ 3.http://www.webmetricsguru.com/category/socialmarketing-analytics/ - there are actually 8 posts I wrote on the subject, not all show up with this URL. 16.http://management.about.com/od/businessstrategy/a/
pg 13
9.http://www.webanalyticsassociation.org/resource/ resmgr/PDF_standards/WebAnalyticsDefinitions.pdf
Links
reorganization 12.http://www.cipr.co.uk/content/cipr-social-mediameasurement-group-launches 13. Phillip Sheldrake on Twitter @sheldrake 14.http://www.webmetricsguru.com/category/socialmedia-case-study-havana-central/ The author was often frustrated why we could not go further here. 15.http://havanacentral.com/vip.php
DataInteg02.htm 17.http://techcrunch.com/2010/06/03/location-2012/ Robert Scoble argues massive amounts of Geo-Location data is being collected in Silos when it would be better for consumers the data was shared across all vendors. 18.http://analyticsinsightblog.stratigent.com/2010/09/ centralize-your-data-using-adobe.html The article suggests Web Analytics platform Site Catalyst is one place where it’s possible to pull together data from disparate sources into actionable dashboards and reports. 19.http://semphonic.blogs.com/semangel/2010/06/ further-thoughts-on-data-warehousing.html - Gary Angel suggest 3 dimensional arrays open up new questions, targeting opportunities, and analysis methods that otherwise don’t exist. 20.http://www.omniture.com/en/products/open_ business_analytics_platform Adobe Omniture Open Business Analytics Platform. 21.http://www.slideshare.net/webanalis...erlands-amsterdam-june-2nd-2010judah-phillips 22. SiteWrench CMS – this CMS seems to have it all.
pg 14
doc_607094351.pdf
This is a report about Spectrum Analytics on Tracking Social Media ROI using.
Tracking Social Media ROI Viewing Spectrum Analytics
Marshall Sponder – Webmetricsguru.com
[email protected] @webmetricsguru 09/05/2010
pg 1
An Introduction to Spectrum Analytics for Social Media ROI
Social Media is a fairly new communications medium and activity that is often considered to be “free” and available to everyone. With abundant free and low cost tools and platforms to create, promote, monitor and analyze social media content businesses and individuals, excited about what Social Media can do for them, too often assume that measurement of Social Media can be taken care of as more of an afterthought. In the past estimating Social Media ROI was acceptable, increasingly that is no longer the case as outlined by TheBrandBuilder, Oliver Blanchard1. As CMOs and CEOs have started to seriously consider Social Media as an effective marketing medium, they also seeking a similar level of precision in Return on Investment (ROI) metrics they have gotten from Direct Marketing. Everyone is attempting to devise a framework for their company’s Social Media ROI, but the challenge is somewhat different for each business and industry. What are the metrics; beyond the basics such as how many Fans and Followers an account has that can actually help management manage better? And, of course, what metrics affect the HOLY GRAIL of business metrics: the bottom line? Lately, a lot of attention has been placed of formulas to calculate Social Media ROI. In fact, earlier this year Altimeter Group and Web Analytics Demystified shared a framework for Social Media Metrics and ROI2 addressing the Marketing Industry Challenge:”I can’t measure social media ROI.” The metrics were generalized to fit into a framework, which I analyzed3 in a series of posts at Webmetricsguru.com. I believe the framework is a step in the right direction, but much more guidance is needed. I will even go further by suggesting we avoid talking about
two dimensional formulas in order to calculate Social Media ROI, even though it is tempting to do so, before we are able to capture the information to put into those formulas. While we can reduce a business’s financials to a series of formulas that calculate profitability or loss, which is fairly well instrumented by accounting software and CPA’s and best practices have been for hundreds of years (for the sake of investment and profitability), it’s much more difficult to do the same thing for Social Media ROI (which is only a few years old).
Why I wrote this paper, finding Ultraviolet Data Nuggets and Rainbow Analytics
Over the past year I worked with Cecilia Pineda Feret4 analyzing the web site and social media of the Havana Central5 restaurant group with a few locations in Manhattan. We attempted to provide ROI information to the owner of the group and we regularly discussed the challenge of tracking our efforts for the company, the brand and the bottom line. For a small business, tracking and measuring this involves looking under the radar and following over the rainbow to find the ROI. Yet, every business has a different pot of golden data nuggets (or rainbow pancakes, they’re easier to find!), or RAINBOW ROI (Cecilia Pineda Feret coined this term in our discussions about Havana Central ROI tracking and analytics). The usual data on the radar are easy to capture: hits, followers, page views, etc. These metrics are more diagnostic according to Stratigent’s Jennifer Veesenmeyer6, allowing you to compare data points over time, and against competitors, and so on without necessarily providing any actionable insight into what the business should change or keep doing. Additional metrics are also required such as knowing results of efforts which were the most effective in bringing in sales, expanding the
pg 1
customer base, and increasing exposure of the brand. Acting as an analyst here my insights are usually connected with data I touch. I reasoned data to prove Social Media outreach is effective in driving new business and revenue was available, but we were not able to capture it. I termed such data as “ultraviolet”, because we can’t see ultraviolet light, yet it’s present all around us. I thought about what it takes to capture the information and tabulate it. When I started this line of thought my goal was not necessarily to solve Havana Central’s problem as much as to prove community management was an effective income generator for any restaurant chain. I decided to create an approach anyone could use to find how much data they were missing and what they could do to enable data capture in preparation for Social Media ROI based on the need to analyze the full spectrum of data and initial discussion of the concept can be found on my blog7. While coming up with this approach we realized Havana Central really needed a connection from the business and how it operates to the customer, and back to the business for tracking via analytics using tactics such as couponing, secret passwords and perks for VIPs, etc. Read on to find out more about the methodology. Similar diagnostics can be obtained from Google Analytics Benchmarking (when sharing data with Google) based on business type. The charts below show that on average, Havana Central gets 3.75 times more visits than the average restaurant using Google Analytics but the average restaurant is generating more pageviews per visit (it may have more to do, in this case, with PDF menus that are downloaded but not counted as pageviews or otherwise tracked by Google Analytics unless virtual pageviews are created for each menu). Havana Central’s Referral traffic as detailed by Compete. com shows no traffic at all from Social Networks (low sample site) and so a marketing goal ought to include meeting or surpassing the average referral traffic from Social Networks for the relevant category – in this case, several hundred more visits per month from Facebook alone would be called for (note: Google Analytics shows some Facebook and Twitter traffic to havanacentral.com). Figure 1 – Compete PRO Referral Analytics Traffic Dashboard, July 2010 for Industry Category: Entertainment Guide: Restaurants and Dining8
Setting Measurement Goals
Every business falls into some type of category. In order to know what is realistic to strive for in relation to Social Media, it helps to know what the average traffic is and sites for the category you business is in. For restaurants, on average 5% of referral traffic to top restaurants comes from Facebook, 0.16% from Twitter and .27% comes from MySpace.com according to Compete’s Referral Analytics Traffic Dashboard for the Entertainment Guide: Restaurant and Dining Category. (a new offering of Compete I provided suggestions about early on).
pg 2
for VentureBeat /SocialBeat last year10 suggested Video publishers had the IAB’s VAST standard to thank for rapid monetization of video ads that started taking place; I quote from my article here: … Monetizing online video has been somewhat problematic because the technology in place interfered Figure 2 – Google Analytics Benchmarking of Havanacentral.com vs. all other restaurant sites in Google’s database. But that’s beginning to change, partly due to the IAB Using benchmarking tools like Compete, Quantcast, Comscore, Nielsen, Alexa and in this respect, Google Analytics, is useful in order to formulate some business goals (which can also turn into measurement goals), especially those connected to Social Media, but they are not a replacement for direct measurement using site analytics. In fact, Web Analytics data and Compete.com data almost never match as the methods to obtain the information in both cases are different which accounts for the disparity in numbers of unique visitors, visits, pageviews, etc. In my discussions with Compete, they have also pointed out their tools are meant to complement, not replace site analytics. It’s also notable Comscore, Compete and Quantcast allow JavaScript tags to be added to your website which allow more accurate measurements for you business (but not necessarily your competitors unless they are also adding JavaScript tags to their sites). My work at the Web Analytics Association as Board Director in 2007-2009 led to creating and leading the Social Media Committee (now-defunct)11; it showed me the industry lacked consensus on social media standards and many of the players I talked to at the time liked it that The Web Analytics Association released Web Analytics Standards a few years back and there have been similar
9
with easy communications between publishers and third party video ad platforms.
(Internet Advertising Bureau), via the VAST Standard. According to the IAB, VAST (Video Ad Serving Template) includes a standard XML-based ad response for in-stream video and an XML Schema Definition (“XSD”) for developers. It’s meant to accommodate the majority of current practices within the online digital video advertising business. The gains we‘re seeing now were almost impossible to achieve until the VAST Standard, which made it easier to obtain end-to-end control between ad servers and associated video players (said CEO Tod Sacerdoti) Because we lack standard definitions for influence, conversations, engagement and conversions, to name a few, it’s so very much harder to reach an interoperable place where all the platforms that need to talk to each other, can, in a meaningful way.
Missing Measurement Standards
way (standards may be seen by some as a way to stifle innovation). I don’t know if that’s still true. On the other hand, for Social Media to “grow up” Standards definitions may be needed; I for one, believe they are very needed and important work is being done
pg 3
attempts by the Internet Advertising Bureau for Banner Advertising, Social Ads and Video Ads. Setting up standards may be a necessary step before widespread monetization can take place. In fact, an article I wrote
in the UK to achieve those standards through the CIPR (Certified Institute of Public Relations) Social Media Monitoring study group and panel , of which I’m an
12
person do Community Outreach, Management and track catering orders in a spreadsheet that could work for a very small operation that can live with a manual process but quickly becomes un-scalable. It also brings up another issue; at what point does tracking and monitoring roles interfere with going out and generating business? Marketing tasks are often assigned to different people/groups for reason; unless those people are the same page in every respect that matters how well can we expect them to communicate, or even agree on what needs to be done, needs to be tracked and how? The communications process is often too taxing for many
honorary member. I hope readers will follow CIPR-CP leader Philip Sheldrake quest to come up with standards
13
for Influence. In this paper the lack of standards could mean any measures a company comes up with to measure its own social media efforts (hopefully after aligning their business and measurement goals and strategies) will be difficult to apply universally, outside the context a specific business that came up with their own measures. That is OK for the short term, but in the longer term, we need consensus on what we measure and how to put that measurement in place. Once we have that, Rainbow Analytics will be much easier to achieve.
SMB’s and it’s difficult even for large corporations but larger businesses are usually willing to make strategic investments in messaging technology, business and social media measurement processes while SMB’s, often just trying to survive in difficult economic times, cannot or are unwilling to try. It might seem that communications problems (such as two individuals being on the same page) would be easy to solve, but it’s usually not the case. Another example from my experience working with Havana Central was creating and fine tuning Radian6 Alerts in order to know when customers were in a location and offering them a drink14 (or a coupon, which was never set up). I saw the potential here but got frustrated with the results. You can’t blame it on software, automation (Radian6, FourSquare and Facebook Places to name a few) took care of knowing people are in a venue by sending us email alerts (a measurement goal for Social Media attribution); but the business goal of Brand Awareness and Customer Loyalty (which are much harder to align and scorecard) was not aligned with the measurement strategy I came up with (getting management to be on the same page with business and measurement strategies is the subject of several books onto itself and I’m not going over it here).
pg 4
Aligning Business Processes to Measurement Goals
A common challenge for Social Media Measurement and the biggest roadblock finding the data nuggets under the radar and over the rainbow are business processes that are unaligned with a business’s goals for measurement. One example, obtaining Requests for a Quote (for a potential catering) which corresponds to a page on the Havana Central site (monitored by Google Analytics on one side and back end email request on the other) typically require communications between a marketing assistant and community manager (communications didn’t happen often enough or were too difficult to maintain). Catering jobs generate several thousand dollars of income, per job, to the bottom line, but the business process wasn’t yet aligned to the measurement goal (attribution of a referral to outreach). One solution is to have the same
rest of the organization). At least, that’s how it looked to Testing this approach in the real world of a NYC restaurant chain we found no systems in place yet to incentive very busy struggling restaurant managers, waiters and waitress (who need to be trained and often change jobs) to gather the information (a customer is in a location, greet them and offer a coupon) and record it in such a way that it can be attributed to Social Media and then tabulate into an ROI metric. Specific middleware to record the transactions was also missing. This lack of alignment is the biggest problem we have as the tools are in place to measure most of what we need already and plenty of people have come up with measures for Social Media ROI, but there’s almost succeeding in defining and aligning the social media dimension of your business goals to your overall measurement goals leading to the common result of Social Media ROI that cannot be accurately defined or measured (but is sometimes estimated); several people are working on this problem right now but no one I know of has solved it. Everyone reading this paper can likely attest to many instances of the very same alignment issues in their own organizations, once you step back and think about it. There may be several valid reasons why your business strategies aren’t aligned with measurement strategies; for example, when I worked at IBM.com (2003-2008) we were usually unable to track marketing and search optimization activities all the way down the sales channel. Part of the problem was traced back to technology (accounting systems that don’t speak to each other), another part were was sheer volume of activities and people (finding the right people in each part of IBM to talk to each other was problem) but the biggest issue did not turn out to be technology, as much as it was business processes for sales organization appeared to be different, and in conflict, with the measurement goals of the dotcom (IBM.com operates as a separate business unit from the
pg 5
me. Each part of an organization has its own goals and methodologies to serve its own purposes and to get work done; weather those purposes support or interfere with accurate measurement is up for you, the reader, to decide. In fact, just about every business unit, every company, every corporation, every city, state and every country have the same exact communication issues around goals and strategies, and as we have seen throughout our civilization’s history, this is not an easy problem to solve. Maybe the best way to look at our differences as stepping stone to better way of communication that serve us all (once there is enough consensus and people in place at the top that can focus business and measurement goals, strategies and policies).
Figure 3 – Radian6 charts showing Social Media Buzz Measurement of Havana Central Times Square location vs. neighboring restaurants Blue Fin, Heartland Brewery, Carmine’s Times Square, and Bond 45. The chart on the right shows the number of twitter followers of those who tweeted about each restaurant in the last 3 months and where they are located by region. Note: this chart shows aspects of Brand Awareness for the 5 restaurants surveyed here.
Why Social CRM is needed
Over the last year Social CRM has quickly evolved as means to connect Social Media Chatter to existing records in Salesforce and other CRMs. This space is quickly becoming one of the hot topics in Social Media Monitoring and will be subject of a future paper by the author.
Finding the right middleware to record results
This lack of alignment is the biggest problem we have as the tools are in place to measure most of what we need already and plenty of people have come up with measures for Social Media ROI, but there’s almost no one succeeding in defining and aligning the social media dimension of your business goals to your overall measurement goals leading to a common result of Social Media ROI that cannot be accurately defined or measured (but is sometimes estimated); several people are working on this problem right now but no one I know of has solved it.
Closing the Gap and Finding the Rainbow
Data can be collected from Social Media Monitoring platforms like Radian6 easily enough, but connecting a visitor (a social media awareness goal & a measurement goal) with increased sales by restaurant location generated from social media customers as a result of social media outreach (a business goal) takes a level of middleware (Social CRM, covered in the previous section) plus more buy in and patience from management (it’s often a hard combination to come by). Also needed is a way of connecting monitoring platforms such as Radian6 and the backend check receipts; today is only being envisioned and no one that I know of has done it yet. While this paper focuses on restaurants with its focus on location, check-ins, cash receipts, catering and reviews, the same process can be applied to any business with slight modifications. Barcode / QR code readers, RFID and Geo-fencing are expected to close the gap for restaurant venues in the next two years but the software to connect all this data must be
Figure 4 – part of the authors’ visualization of a Radian6 date export of visits by Twitter Accounts to any Havana Central location from June 09-June 10 (anyone checking in via FourSquare, or tweeting they were eating at one of the locations), seehttp://www.webmetricsguru.com/category/ social-media-case-study-havana-central/ for more details.
CRM aware and I expect the role of community manager will evolve into Relationship Management within the next 18 months. Acquisitions by Scout Labs by Lithium Technologies, Radian6 partnerships with Salesforce and recent enhancements by Alterian around the SM2 platform are just the beginning of the tip of the data iceberg lying
pg 6
under the surface, just waiting to be tapped.
example, each tweet about eating at the restaurant can be estimated to be worth over $100 based on the average order size, the average party size and the average price of a meal). Estimating Social Media ROI is useful when you make a
Getting Started by Performing a Social Enablement Audit
The only way to know how much precious data your business has in the unseen spectrum is to audit your data. Specifically, list all the sources of data your business has and what campaigns and marketing initiatives your business is running. This table shows data from Havana Central.
decision to expand or retract a marketing initiative but is not a replacement for actual ROI measures that are tied to a transaction. In another case, traffic and possible marketing efforts directed at the VIP page of the Havana Central restaurant site15 can be expected to appear in Google Analytics and Google Adwords (when ads are run) but not anywhere else.
Figure 6 – How enabled are we for Social Media ROI? Figure 5 – Social Enablement Audit of Havana Central I put all the data sources vertically and all the campaigns, including website promos and pages horizontally. Note, the x- and y-axis could be reversed easily. By conducting an audit of all the available data sources within in the context of all the known marketing campaigns and initiatives, we learn what information we have, where to find it, what format it is in and the level of difficulty in overlaying data in order to create Social Media ROI metrics we need. For Social Media ROI calculations that are tied to an actual In this example, data validating community management shows up partly in Google Analytics and Radian6, but will fail to appear anywhere else. It is next to impossible without a guaranteed tie into a revamped business process to track the actual dollar value to the outreach being done in behalf of the business, though one could estimate it (for transaction that takes place at the restaurant or its online site we need a logical way to overlay data, using what is commonly called a common key18, that currently doesn’t exist in this business context. Figure 6 suggests, based on the Enablement Audit that 82% of the data is “ultraviolet” or invisible (or, not able to be combined because a
pg 7
Most data is collected in Silos16, it’s pretty much always been that way and is getting worse according to Robert Scoble17. Applications are designed to provide specific functionality and data are collected around it, but usually applications aren’t designed to integrate with anything else – in this case, we don’t expect a reservation site such as Opentable, handling most restaurant reservations to have data on Community Management outreach of the VIP page of a website.
common key does not exist). One approach used by large corporations is to overlay data into Data Cubes or data warehouse where all the data you have can be imported into a Cube19, but usually the resources to pull something like this together are beyond reach of most businesses, including those who want measure ROI the most, small businesses and even PR Companies doing social media programs as part of Public Relations and Marketing (in behalf of their clients). Brick and mortar SMBs who could not possibly conceive, afford or execute such a data cube even though they may actually require one, but once your data is in the Cube you can perform analysis that could uncover ROI and customer behavior that is not ordinarily visible. Figure 8 – Data Coverage In the process of doing a Social Enablement Audit you might find that you are not collecting all the data you needed; in this case you may not have some data that you need in any form and in any data source. The best way to find out if you missing collecting data is to compare your data sources to each other. As you examine your data you may find there are some things you want to know that no application is tracking at all and this will drive new solutions for your business. What we’re trying to find out here is how compatible is the data being used to run your business to each other and how easy is it to integrate information. Typically, applications are not created to integrate and the data within each source and Figure 7 – slices of a data cube Given the lack of the needed common key (in this case, a common record locator for every record in Google Analytics, Opentable, SeemlessWeb, Email Marketing, etc) in many data sources – information needed to populate the data cube is going to be spotty, noisy and in other cases hard to translate it to the right structure to be effective. According to Gary Angel, the CTO of Semphonic, when you integrate data into a data warehouse, you open up new questions, targeting opportunities, and analysis methods that otherwise doesn’t exist19. are frequently incompatible with each other. One attempt to answer the compatibility problem was Adobe (formally Omniture) Open Business Analytics Platform20 using Web Analytics as central point to unify disparate applications such as Radian6 (one of our data sources), several different Content Management Systems, Salesforce Customer Relationship Management (CRM) and even Opentable, using Omniture for their own marketing analysis, (using Test and Target) within Site Catalyst, but all of these require a lot of money and time that do not make sense to implement for most small business (not until someone comes up with a suite of tools for low end of the market that makes data integrations much easier and inexpensive).
pg 8
Supposing you could integrate all the data you need into a cube, would the resulting reports lead to insights not ordinarily obtainable? I think so, but it will depend on how well business processes are translated in to the matching reports via a clearly defined measurement strategy, mentioned earlier, and gifted analysts needed to make sure they do. Hopefully, business will also learn how to listen to what the analysts have to say when they are able to come up with marketing insights and give them a seat at the table which they traditionally haven’t had in most organizations. I believe it’s time to consider doing similar enablement for Social Media that we more typically perform for Web Analytics (with extensive page tagging often required to pick up specific customer behavior, often a measurement goal).
could use the data to analyze their sites, but most businesses didn’t know how to define their business goals and translate them into measurement goals and align their organizations so that real analysis could happen with the result good analysis was done, but rarely. Around 2005 things got to be more interesting with a convergence of sorts including Flash and Ajax tracking evolving, but again, most organizations aren’t yet able, willing or think it’s too expensive to translate what they do and what they want to track into what these tools can do and what they can track, and as a result much data and insight is lost. As social media began as grassroots, peer to peer expression of communications between friends and fans, etc, this kind of integration had not lent itself to being instrumented in the same way Site Analytics evolved but it’s only a matter of time until it does get instrumented, much as Web Analytics has. As corporations are going more social, so are SMB’s. Parallels between Social Media Measurement and Web Analytics suggest convergence and the possibility of actual ROI measurement once business goals are merged with right measurement goals. With geo-location coming into the mix and checking in quickly becoming the universal currency, it may become easier to integrate social media with the rest of the data
Figure 9 – Stages of Web Analytics Evolution from Judah Phillips deck on Webanalytics congress the Netherlands, Amsterdam June 2nd 2010, slide 14 .
21
sources – and the cost and complexity will decrease on the low end, making possible for SMB’s to have a plug and play solution within the next two years. However, translating your business process, connecting the Data Rainbow back to measurement will be required even in streamlined solutions, leaving room for a lot of creativity.
In the beginning of Web Analytics circa early 1990’s basic diagnostic data is all that could be collected and today, many businesses are capable of tracking much more than this (with proper enablement) but settle for using the basic gauges of visitors, visits, pageviews though little insight into business can be obtained this way. At the turn of the century analytics improved and business
Finding Data Nuggets of Gold under Data Rainbow
It’s my belief we do not achieve actual Social Media ROI
pg 9
unless we can segment social media activity and associate it across data sources and marketing channels back to actual revenue. In fact, most business and consultancies have no idea how to start, which is why it doesn’t happen often, if it does at all.
pancakes!)
In order to know what can and can’t be combined to form Social Media ROI we need to look at the data we’re collecting in greater detail – assessing what we actually have to work with.
Figure 11 – A Opentable Report An Opentable report as it is currently generated is almost impossible to attribute to Social Media outreach there’s no common key to associate with a visitor from Facebook (unless the application tracks that). Opentable tracks website visitors that came from Havana Central, but there’s no way to count how much of it was Social Media related in the reports they provide to business owners. The same problem arises with Grubhub, a popular Restaurant Search utility.
Figure 10, Example of a Data Rainbow Only a fraction of the data we need to collect to track Social Media ROI is actually being saved (or seen) in a way that can be analyzed and connect to Social Media Marketing initiatives. The chart above shows some of the individual silos of data for Havana Central (the different data sources above might can be termed silos for the purpose of this paper). Figure 12 – A Grubhub report The Grubhub report has no common key except a date, everything else is tied back to internal databases that are not accessible (i.e.: ID number), no easy way to tie back to Google Analytics (unless we can put tracking code into a Grubhub instance, which is really hard to do for third party applications like Grubhub and Opentable).
Looking at the Data in our Silos for data nuggets (or rainbow
pg 10
Radian6 reports that monitor Buzz around the restaurant do offer possibilities to tracking what website visitors do on your website through its integration with WebTrends and in your business location via Salesforce.com but few businesses have adopted those integrations to my knowledge. As these Social Media targeted integrations exist at all but are rarely adopted, suggest businesses are too often unaware of the choices they have before them when they formulate their business goals and measurement strategies and often choose unwisely or fail to choose at all. In any case, while your Social Media campaigns might be driving a certain level activity which should be measureable, that same activity is invisible to the reports you can currently pull from Opentable and Grubhub, Google Analytics, Facebook and even Radian6 without deliberate measurement enablement and the alignment of your measurement strategy with you business strategy.
used to link all the transactions together it’s still possible to fill the gaps with third party applications and ingenuity.
Finding Solutions to Bridge the Gaps
Figure 14 – Measurement Solutions by Campaign or Marketing Initiative Taking a closer look at a specific marketing initiative against the sources of data available can drive solutions where missing data can be added in when it’s missing. Look at your campaigns and marketing. Figure out how to add the additional enablement that would allow you to track the Social Media ROI of a particular activity such as somebody downloading a menu if they came from Facebook or Twitter.
Comparing data dimensions and combining them in the Data Rainbow
Bridging the Data Gap Example – Connecting Facebook and Opentable for Data Nuggets
Figure 13 – Data Dimensions comparison Figure 15 – Finding ways to connect Opentable with In the chart above I highlighted in green dimensions from different data sources that might be able to be combined to get that Rainbow view and insights derived. We know by now the vast majority of data being collected cannot be combined based on the example above. While your data will too often be missing that common key that can be You may need to develop or find 3rd party applications capable of bridging the gaps found using this audit process. While we can enable Google Analytics/FB connection –we still need a Content Management System
pg 11
that connects a website, Google Analytics and Opentable and Google Analytics can act as a go between.
2) Most businesses, esp. SMB’s are not yet ready to fully implement the data audit methodology. Someone needs to come along and make the process easier for them, and more streamlined, before this is going to happen on an SMB level. I suspect a streamlined implementation of data audits will differ somewhat based on business verticals. However, within each vertical there will be enough similarities to standardize and streamline (i.e.; most restaurants use Opentable and SeemlessWeb and should be part of the solution for this vertical, but perhaps not for other types of retail establishments like Supermarkets, clothing outlets, Travel Sites, Hotels, etc, which will need different sets of data feeds and metrics). 3) Depending on an actual SMBs’ goals, what you are measuring may be different for another SMB, even in the same or another industry. The challenge here is to capture the data that is not immediately seen or considered and
BTW, in the process of detailing needed capabilities our audit concludes the best enablement would be rebuilding that website (ouch!) with a CMS (SiteWrench CMS) that can handle Opentable & Google Analytics integration22.
relatable back to a Social Media effort by linking your measurement strategy and goals with your business strategy and goals. 4) To capture this data, we need to perform a complete audit of efforts made. It may be a time consuming task depending on the size and complexity of the business, but for the moment remains the best way to fully consider and fully integrate efforts across multiple platforms. 5) The best place to integrate all the data sources for
Conclusions:
1) “States of Readiness” for Social Media ROI Analytics implementation are very similar to those for Web Analytics, where there are various levels of participation using the tools currently available. For example, Radian6, Salesforce CRM and WebTrends integrate but most businesses are not ready to take advantage of it yet. Once business moves from past collecting basic diagnostic data, they can begin to merge their business strategy and goals with specific measurement strategies and goals and the measurement of Real Social Media ROI can begin.
your business is within the analytics platforms such as Google Analytics, Adobe Site Catalyst (formerly known as Omniture) though other solutions such as Spredfast are available. 6) You need to work with your web developers, programmers, product and sales teams to enable tracking of certain actions to ensure that your efforts are being tracked and what meaning you assign to them other than just collecting them on a periodic basis like many continue to do.
pg 12
4.http://www.linkedin.com/in/ceciliapinedaferet Cecilia Pineda Feret’s LinkedIn page. 5.http://www.havanacentral.com Havana Central Website 6.http://www.stratigent.com/news-and-events/pimp_ your_reports/default.html Stratigent’s Pimp Your Reports program which I recently attended 7.http://www.webmetricsguru.com/archives/2010/08/ on-360-degree-analytics/ - put forward that most of the interesting data that proves Social Media is working is invisible to the naked eye – or “ultraviolet”. However, with the right devices and mythology the data can be seen and quantified 8.http://www.compete.com/referrals/referrals/ category/3037/?threshold=1 What you’ll find at the end of the Data Rainbow …… Yum! (Amanda -http://iammommy.typepad.com/my_ weblog/2009/10/rainbow-pancake-sugar-cookies-winniethe-pooh.html) 10.http://social.venturebeat.com/2009/11/29/brightrollsays-video-ad-profits-are-soaring/ 11.http://www.webanalyticsassociation.org/?c_ 1.http://thebrandbuilder.wordpress.com/2009/05/29/howto-not-calculate-social-media-r-o-i/ Oliver Blanchard has written several posts and a popular SlideShare presenting his views on what is and is not Social Media ROI. 2. Altimeter Report: Social Marketing Analytics (Altimeter Group & Web Analytics Demystified) seehttp://www.webstrategist.com/blog/2...ing-analytics-with-web-analytics-demystified/ 3.http://www.webmetricsguru.com/category/socialmarketing-analytics/ - there are actually 8 posts I wrote on the subject, not all show up with this URL. 16.http://management.about.com/od/businessstrategy/a/
pg 13
9.http://www.webanalyticsassociation.org/resource/ resmgr/PDF_standards/WebAnalyticsDefinitions.pdf
Links
reorganization 12.http://www.cipr.co.uk/content/cipr-social-mediameasurement-group-launches 13. Phillip Sheldrake on Twitter @sheldrake 14.http://www.webmetricsguru.com/category/socialmedia-case-study-havana-central/ The author was often frustrated why we could not go further here. 15.http://havanacentral.com/vip.php
DataInteg02.htm 17.http://techcrunch.com/2010/06/03/location-2012/ Robert Scoble argues massive amounts of Geo-Location data is being collected in Silos when it would be better for consumers the data was shared across all vendors. 18.http://analyticsinsightblog.stratigent.com/2010/09/ centralize-your-data-using-adobe.html The article suggests Web Analytics platform Site Catalyst is one place where it’s possible to pull together data from disparate sources into actionable dashboards and reports. 19.http://semphonic.blogs.com/semangel/2010/06/ further-thoughts-on-data-warehousing.html - Gary Angel suggest 3 dimensional arrays open up new questions, targeting opportunities, and analysis methods that otherwise don’t exist. 20.http://www.omniture.com/en/products/open_ business_analytics_platform Adobe Omniture Open Business Analytics Platform. 21.http://www.slideshare.net/webanalis...erlands-amsterdam-june-2nd-2010judah-phillips 22. SiteWrench CMS – this CMS seems to have it all.
pg 14
doc_607094351.pdf