Octavianworld

Controlling for Impression Volatility in Digital Ad Spend Tests @DataXu

I've recently been involved in evaluating the results of a matched market test that looked at the impact of changes in digital advertising spend by comparing test vs. control markets, and by comparing differential lift in these markets over prior periods (e.g., year on year).  One of the challenges involved in such tests is significant "impression volatility" across time periods -- basically, each dollar can buy you very different volumes of impressions from year to year.  

You can unpack this volatility into at least three components:  

• changes in overall macro-economic conditions that drive target audiences' attention,
• changes in the buying approach you took / networks you bought through, due to network-specific structural (like what publishers are included) and supply-demand drivers (like the relative effectiveness of the network's targeting approach)
• changes in "buy-specific" parameters (like audiences and palcements sought).  

Let's assume that you handle the first with your test / control market structure.  Let's also assume that the third is to be held constant as much as possible, for the purposes of the test (that is, buying the same properties / audiences, and using the same ad positions / placements for the tests).   So my question was, how much volatility does the second factor contribute, and what can be done to control for that in a test?

Surfing around I came on DataXu's March 2011 Market Pulse study.  DataXu is a service that allows you to buy across networks more efficiently in real time, sort of like what Kayak would be to travel if it were a fully automated agent and you flew every day.  The firm noted a year-on-year drop in average daily CPM volatility from 102% to 42% from May 2010 to February 2011 (meaning I think the average day to day change in price across all networks in each of the two months compared).  They attributed this to "dramatically increased volume of impressions bought and sold as well as maturation of trading systems".  Notwithstanding, the study still pointed to a 342% difference in average indexed CPMs across networks during February 2011.  

A number this big naturally piqued my interest, and so I read into the report to understand it better.  The top of page 2 of the report summary presents a nice graph that shows average monthly indexed CPMs across 11 networks, and indeed shows the difference between the highest-priced and the lowest-priced network to be 342%.  Applying "Olympic scoring" (tossing out highest- and lowest-priced exchanges) cuts that difference to about 180%, or roughly by half -- still a significant discrepancy of course.  Looking further, one standard deviation in the whole sample (including the top and bottom values) is about 44%.  Again, though perhaps a bit less dramatic for marketers' tastes, still lots.

(It's hard to know how "equivalent" the buys compared were, in terms of volumes, contextual consistency, and audience consistency, since the summary doesn't address these.  But let's assume they were, roughly.)

So what? If your (display) ad buys are not so property-specific / audience-targeted that run-of-network buys in contextual or audience categories are OK, future tests might channel buys through services like DataXu and declare the buys "fully-price-optimized" across the periods and markets compared, allowing you to ignore +/- ~50% "impression volatility" swings, assuming the Feb 2011 spreads hold.

However, if what you're buying is very specific -- and only available through direct purchase, or one or two specialized networks at most -- then you ignore factor 2, trust the laws of supply and demand, and assume that you've bought essentially the same "attention" regardless of the difference in impressions.

I've asked some knowledgeable friends to suggest some perspectives on this, and will pass along their ideas.  Other feedback welcome, especially from digital advertising / testing pros!  Oh and if you're really interested, check out the DataXu TC50 2009 pitch video.

Caveman BI

Many client project workplans I intersect have  the line item "Determine KPIs".  Without further context to frame this work, folks typically end up with something that:

• is (too) narrow, functionally
• is impermanent -- KPI reporting and analysis lasts only as long as the effort itself is sustained
• is disconnected from / uncoordinated with other, interdependent parts of the business

Now, consider the view from the analytic teams supporting efforts like these.  When multiple such requests come in, it's left to these teams to prioritize them, and reconcile them semantically and logistically.  This is hard.

To deal with these challenges, many BI (Business Intelligence) tool vendors provide pre-built "business models" that suggest a set of KPIs based on a more fundamental concept of the operations they reflect.  These can be useful, but they're not free, and it's a lot of work to retro-fit them to your business.  This is expensive.

Whether you use a vendor model as a point of departure, or roll your own, I've found it useful to have a basic "Caveman BI Model" that serves three purposes:

• it provides a common language to the folks involved in the KPI development effort
• it provides a broader context on how KPIs fit into the business, so folks don't lose the forest for the leaves
• it starts simple, so your team can work its way up to the complexity you need, rather than trimming back from the complexity it's given by a pre-existing model

The Caveman BI Model starts with a plain-English, noun-verb description of what the business does that needs to be understood better:

"We use resources to create products that we sell to customers with content through channels."

Generic and esoteric maybe, but in one sentence we've described the entities of the business that we need to understand better:

"We use resources to create products that we sell to customers with content through channels."

Put another way, this one sentence describes "things I need to know about to run the business".

Next, we use metrics to describe how (well) we're managing the entities.  Metrics have the following progression of development:

• a raw count: "we have n customers"
• a synthetic measure, commonly a quotient, produced from more than one count: "$x in sales for #y in customers"
• an index that allows us to compare a raw count or a quotient longitudinally (against past performance), cross-sectionally (say, against competitors or industry averages), or as a variance vs. a budget / plan / forecast

Now, each of these entities has dimensions that allow us to "slice" metrics by to dig into the business a bit.  For example:

• resources would include capital, people, raw materials, and fixed assets, each of which might have their own characteristics (equity vs. debt; hourly vs. salaried; source; etc.)
• products could be characterized according to needs served (say, a disease condition for a pharmaceutical), or an underlying technology or legal framework that makes it possible (say, "Forty-Act" in the mutual fund business)
• customers can have a number of properties that allow us to segment them usefully; these might include wealth, how big their relationship with us is, geography, gender, age, etc.
• content would include "campaigns" that use different marketing assets, themes, or promotions
• channels covers distinctions like "direct sales" vs. "distributors", or "TV" vs. "digital"

(Oversimplifying: Ralph Kimball's BI bible, The Data Warehouse Toolkit, calls metrics "facts" and suggests constructs called "fact tables" and "dimension tables" that relate facts and dimensions to each other.  Put another way, what are the different dimensions that we have to slice each fact by in order to understand the business?  "Sales to customers by product and customer geography" would be an example of an intersection we might want to know about.)

These dimensions in turn have hierarchies.  Geography, for example, has "city, state, country".  Some hierarchies are helpfully scalar:  second-minute-hour-day, for example.  But, hierarchies are often a big problem because they don't reconcile neatly.  For example, weeks don't divide neatly into months; some industry data may be sourced from places that use different hierarchies, making apples to apples comparisons hard.

That's pretty much it.  Summing up:

• what do we want to know about? entities
• how do we want to measure it? metrics
• how do we need to slice this measurement? dimensions
• how do we roll up / roll down these slices? hierarchies

So how do we define and distinguish KPIs  from the myriad metrics we could choose?  One rule of thumb I've applied is to choose metrics that start and end a process.  For example, "qualified leads" are an output of a marketing process that become an input to a sales process.  A transaction is then the end of the sales process.  Each of these could be a KPI; a "synthetic" KPI might compare leads to sales as a "conversion rate".

If you're charged with coming up with KPIs and then provisioning them for decision-makers (who may include you too!), I'd suggest you start by asking someone to explain what the "Caveman BI Model" is for your business, using this framework as your interview guide.  If you find that such a model has not been expressed per se, or doesn't exist, take a pass through creating one, maybe by trying to tie some of the key numbers you see used a lot back into the framework.

Again, the purpose of this post isn't to be the last word in Business Modeling, but rather to share a basic approach you can carry around in your head -- and in common across your heads -- that's been useful to me in the past.  Please let me know what you think!

 

Analyzing #Analytics

Take a look at this chart:

Here's what I see:

• interest bumping along to '06
• growing interest through '08 (Competing on Analytics appears March 2007)
• hype-y growth in interest through '10
• wildly uneven interest so far this year
• growing interest in the early part of each year
• summer "blahs" (Who Googles "analytics" on the beach? Don't answer that...)
• bump in interest in the fall, when business plans featuring "analytics" are developed and presented
• quiet Decembers

What do you see?  Are we watching "analytics" jump the shark?  More likely, we've chowed down on it as much as we can, and are now digesting.  Further interest will likely be less at the general conceptual level, and more at the "how do I get there" level.  Hence "Pragmalytics" (Part II here).

Stay tuned for more.

The Marketing Funnel is a Brita Water Filter #OMMAMetrics @judah @schmults

Something you hear a lot these days is, "The 'Marketing Funnel' concept's dead.  It's just not clear what's replaced it."  At the recent OMMA Metrics conference in NY, IBM/Unica's Yuchun Lee described its successor as some sort of "spaghetti".  Judah Phillips had an excellent article in yesterday's Online Metrics Insider (Thanks to Rob Schmults for pointing me to it!) in which he suggested a "tumbler" metaphor and "seeking-shopping-sharing" structure for what we now do.

Here's my entry into the rename-the-funnel sweepstakes: The Brita Water Filter.

Let's consider our requirements for a metaphor to succeed the "funnel":

1. We still have a "current" whose power combines the "push" of customer needs and desires with the "pull" of companies with products that could satisfy those. ("I have testosterone, Porsche makes 911s.")  To me, that still makes "linear" metaphors useful.

2. For me, "Attract", "Engage", and "Convert", plus sometimes "Retain" -- or variants thereof, like Judah's, still work for me as basic stage descriptors.  What's changed is that channels have exploded in number, and audiences have fragmented as they use different ones.  So a good metaphor will describe a journey that is less predictable in both flow path and rate. (It's probably also useful to use stage descriptors that reflect the customer's perspective, not the marketer's -- "Awareness", "Consideration", "Purchase", plus sometimes hopefully "Loyalty" and  "Advocacy" -- but they're not quite as punchy.  Nonetheless, you get the point.)

3. The channel system that lies between nascent demand and final purchase is, these days, much more replete with advice that educates folks as they flow through, qualifying and intensifying final demand.  The metaphor we use has to describe the "chemistry" that happens in these intermediate spaces and interactions.  And, the behaviors we observe in these stages should inform, as Judah suggests in his article, how we market afterward.

4. Over time, however, the system in between gets clogged with a lot of junky information, or becomes technologically obsolete, so you need to refresh or replace your presences in this system on a regular basis.  The metaphor has to anticipate this need as well.

5. The operational and analytic processes for marketing within this flow are higher-tech than a simple funnel.

OK, so let's look at the Brita filter:

1. Water still generally flows through in a linear fashion; and, not all of it flows through at once.

2. As water flows through the charcoal however, it breaks up into much smaller droplets, that flow through at different rates and and less predictable paths.

3. The interaction of the water with activated charcoal only got rid of some impurities; Brita's engineers learned to add an ionic coating to get some minerals out of the water too.

4. Ideally, you change it now and then to keep its performance up.

5. It's higher-tech than a funnel (and more expensive too, but the benefits are sometimes worth it).

So, you say, "Cute but esoteric -- how is this conceptualization useful?"

1. For me, it helps me to validate frameworks I'm familiar with, like the Channel Pathways (TM) framework we used years ago at Monitor / Marketspace, or the similar "Wiggly Line" chart Akin Arikan and Don Peppers describe in Multichannel Marketing.  

2. It also helps me to extend them, by considering differential flow rates through discrete paths, and to focus on what we can learn from interactions at one stage and channel that might inform what we do in subsequent, different ones.

3. It helps to remind me that when populating and managing any such framework, when it gets too complex or overly-"excepted" with rules, I might be better off replacing it.  

4. It helps me to remember that at best I can only be probabilistic and not deterministic in my understanding of "customer flow dynamics", and that what's important is to be explicit about probability levels so that group decisions can be helped by a shared understanding of these.

5. Finally, it's oddly memorable (to me anyway), and the filter's specific properties help me remember the requirements better.

So, what do you think? Please answer the poll below, and comment with questions / alternatives.

 

Click Here for PollOnline Survey
Conjoint Analysis
| Polls
| Mobile Surveys

| Feedback Tab
View MicroPoll

 

Analytics Commons Post in Google Analytics Blog Today @analyticscommns @linchen @perryhewitt #analytics

Our Analytics Commons project (which I previously wrote about here) got written up in a post on the Google Analytics blog today. ( Thanks to Nick Mihailovski at Google, and to Perry Hewitt at Harvard!  And of course to my partners Lin and Kehan at New Circle Consulting!)

Extending Marketing Integration to Agencies: The "Agency API" @rwlord #rzcs

My friends at Razorfish kindly invited me to their client summit in Boston this week.  It was a great event; they and their clients are working on some pretty cool stuff.  Social is front and center.  Close behind: lots of interesting touch / surface computing innovations (Pranav Mistry from the MIT Media Lab really blew our minds). 

In his opening comments Wednesday, Razorfish CEO Bob Lord made the point that for modern marketing to be effective it's got to be integrated across silos of course; but, further, that this has to extend to having agencies work together effectively on behalf of their clients, as much as clients responsible for different channels and functions need to themselves.

I've been wondering about this as well recently, and Bob's comments prompted me to write up some notes.

One observation is that *if* marketers are addressing agency collaboration, they usually start with an *organizational* solution that brings agencies together from time to time.   While this is great, it's insufficient.  To make entities like these effective, it helps to lay a foundation that *registers* and *reconciles* different aspects of what agencies have and do for their clients.  This foundation, realized through a simple, cheap tool like Basecamp if not the marketer's own intranet system, could include:

• a data registry.  Agencies executing campaigns on behalf of their clients end up controlling data sets (display ad impressions and clicks, email opens, TV ratings, focus group / panel results) that are crucial to understanding the performance of marketing investments, but are typically beyond the scope of what IT's "enterprise data architectures" encompass.  I'm not suggesting that agencies need to ship this data to their clients; rather, that they simply register what's collected, who's got it, and how a client can get it if needed.
• an insight registry.  Agency folks crunch data into fancy powerpoints bearing the insights on which campaign decisions get made.  It would be very helpful if these decks were tagged and linked from the appropriately-permissioned online workspace.
• a campaign registry.  Think http://adverblog.com, only for the marketer's own (and perhaps direct competitors') campaigns of any stripe.  A place to put creative briefs and link to campaign assets / executions that implement them.

These approaches are simple, cheap, and "many hands make light work".  Implementing them collectively as a marketer's "Agency API" would help marketers and agencies to "reconcile" their work, in the following ways:

• discover campaign conflicts and integration opportunities
• unpack possible disagreements into manageable bites to resolve -- data conflicts, insight conflicts, analytic technique conflicts, creative brief conflicts
• better prepare in advance of "interagency council" meetings

Of course the registries are not a panacea for inter-agency "issues" that a marketer needs to sort through, but they help to make any problematic issues more transparent and straightforward to focus on.  

Please take the poll below.  Reactions / suggestions welcome from folks with relevant experience!

Click Here for PollOnline Survey
Conjoint Analysis
| Polls
| Email Marketing

| Feedback Tab
View MicroPoll

 

Analytic Commons Project

With inspiration and encouragement from @perryhewitt, New Circle Consulting and Force Five Partners have launched the Analytics Commons Project (http://analyticscommons.com).  Here's the pitch:

Web analytics is a relatively new field that is evolving very quickly. Fortunately, it's been our experience that the community of web analysts is welcoming, vibrant, and very willing to share. The Web Analytics forum on Yahoo! is a wonderful example of this.

Analytics Commons is an effort to improve on this sharing by structuring it a bit. With structure, we can make relevant knowledge a little easier to find, and we can also make it easier to vet the expertise and reliability of the source of that knowledge. (The new Web Analytics Association Certification program is another good step in this direction.)

In designing Analytics Commons, we also decided to start by focusing on a specific form of analytics knowledge, rather than trying now to architect some general information architecture about the field that could capture all its (quickly changing) variety. In particular, we noticed:

• Google Analytics is ubiquitous.
• We're happy users of it.
• It recently added the ability to share Advanced Segments and Custom Reports.
• While GA has an Apps Gallery that features third-party creations, there is currently no public registry of such shared reports that we're aware of.
• But, there does seem to be pent-up demand for sharing reports.
• And, we had a specific itch we needed to scratch ("Target Towns", more on that below) that would help us Keep It Real.

We also figured we would start with something that would be within our ability to actually get done. Our ambition for this initiative doesn't stop here, however. So, the service also provides a way for visitors and users to suggest feedback to shape the vision and path for getting there.

So how does it work? If we've done our job well, it's hopefully self-evident. You register on the Analytics Commons site, and tell us a little about yourself, ideally through links to places where you keep your description up to date (e.g., LinkedIn, Twitter, etc.). If you've got a report to contribute, you get the url for it by clicking on the "Share button" in Google Analytics' Custom Reports or Advanced Segments sections from a GA profile in which you have access to them. Then, you add the url to our service and tag and describe what you've shared. If you need a report, you search for it on our service. If you find and try a report, all we ask is that you rate and comment on it to tell us how well it matched what you needed. Hopefully, discussions about each report will happen on our service, but if you want to connect privately with a report contributor, we've made room in our registered user profiles for folks to provide contact information if they wish. If you don't find what you were looking for, we let you store the search on our service, and if something matches in the future, we'll send you an email with the search results. If you want, you can subscribe to a weekly email listing new reports that have been added to our service, or get an RSS feed of the same.

The service is free to its users. Our privacy policy is simple: everything here is public, except your registration email if you choose not to share that. We won't share that with anyone, period. If you share a report, we assume you have the authority to do that. If you comment on a report, please be polite and constructive. We reserve the right to moderate comments, and to ban anyone who posts material we deem to be inappropriate or offensive

We saved some space on our pages for advertising / sponsorship, to help cover the server bills. If you're interested, please contact us.

Questions? Suggestions? contact us if you wish at help@analyticscommons.com.

About "Target Towns"

In our work for a client, we observed the following:

• They target wealthy customers.
• Wealth is highly concentrated in the US.
• Wealthy people are highly concentrated in a few towns.

Therefore, we thought it would be useful to track traffic and behavior from these "Target Towns".

We tried to construct an Advanced Segment for "Target Towns" through the GA UI. It didn't appear to support what we had in mind. So we asked for help. Avinash Kaushik, Nick Mihailovski, Judah Phillips, and Justin Cutroni all helped us with a piece of the puzzle (Thank You all!). In the end, the answer turned out that we needed to use the GA API. But the API also had limits on how much information you could hit it with in a single query. So we figured we needed a service that would pass the towns ("Dimensions") about which you wanted information ("Metrics") past the API sequentially, and then would aggregate and present the results in a usable form.

Then we thought: "This is a report many people are likely to need!" So, the "Target Towns" service seemed like it would be a good candidate to help seed our Analytic Commons initiative.

#OMMA Metrics SF 7/21 "Modeling Attribution" Panel: Proposed Questions

Prepping for http://bit.ly/b7KIQz

Here are some discussion questions we're considering.  What's your "keep / change / drop / add" to this list? And, please take the poll at bottom!

1. How do you define "attribution analysis", and in particular, how do you distinguish it from media mix modeling?  Where are the boundaries with "CRM"? Is it just a question of the fineness of the degree to which you can relate media effects to each other? 

2. Everyone's got a poster child example or case study that illustrates the potential of multi-channel attribution analysis.  For each of you, what's your poster child?

3. What are the practical limits of the utility of this analysis, and what's behind those limits?  Put another way, at what point do you see diminishing returns to investments in attribution analysis, and why?

4. Let's talk techniques for a moment, since the session description promised we would.  What particular approaches to data integration or to statistical analysis do you find robust enough  for real-world use, and which ones are still too fragile to make work in most cases?

5. Some people approach attribution analysis as a capability-building exercise -- if we build it, they will come -- while others come at it from a tight, hypothesis-driven focus about where the value might lie, and in what sequence it might best make sense to explore connections.  What do you need to know about a client to suggest which way, or what mix of the two, might make sense for them?

6. What are some principles and tactics for the governance of attribution analysis -- things like reconciling metrics and comp with the global optimization it implies -- that make sense here?  Can you tell any stories about how these have been developed and applied?

7. What kind of experience, training, and topical education does it make sense for marketers to have, to fully take advantage of this?

8. How do you operationalize this analytic capability -- how do you make sure we move beyond insight and apply, either manually or even in automated ways, the directions the analysis suggests?  And, how does this prospective operationalization [warning: mouthful!] affect the analysis you consider doing, in turn?

9. Let's build the ideal RFP for attribution analysis efforts, to help our clients out.  What's in it?  What shouldn't be?

10. What's your recommended reading list for learning about and following this field?

OMMA Metrics & Measurement "Modeling Attribution" Panel SF 7/22: Hope To See You There

I'll be moderating a panel at the OMMA Metrics & Measurement Conference in San Francisco on July 22.  

The topic of the panel is, "Modeling Attribution: Practitioner Perspectives on the Media Mix".  Here's the conference agenda page.

The panel description:

How do you determine the channels that influence offline and online behavior and marketing performance?  

How should you allocate your budget across CRM emails, display ads, print advertising, television and radio commercials, direct mail, and other marketing sources? 

What models, techniques, and technologies should you use develop attribution and predictive models that can drive your business? 

Do you need SAS, SPSS, and a PhD in Statistics? 

Does first click, last click, direct, indirect, or appropriate attribution matter – which is best?

What about multiple logistic regression? 

What is the impact of survey and voice-of-the-customer data on attribution? 

Hear from experts who have to answer these questions and tackle these tough issues as they work hard in the field every day for their consultancies, agencies, and brands.

So far, Manu Mathew, CEO from VisualIQ, and Todd Cunningham, SVP Research at MTV Networks, will be participating on the panel as well.

Hope to see you there.  Meanwhile, please suggest questions you'd like to ask the panelists by commenting here.  Thanks!

MITX Panel: "Integrating Cross-Channel Customer Experiences" (April 29, 2010 8-10a) Part II

We've assembled a terrific panel for tomorrow's event:

• SmartDestinations' Rob Schmults is also a Creative Good Council Leader;
• Judah Phillips is a leader at the cutting edge of analytics in his role at Monster;
• At Staples, Colin Hynes plays a leading role in figuring out store / digital integration, and is heavily focused on mobile's role in that;
• VisualIQ's Manu Mathew sees a broad assortment of situations in facilitating his customers' efforts to develop a cross-channel perspective and optimize based on it.

Here are some of the questions we thought to cover:

• What integrated experiences do you look to as best practice models?
• What are you doing in your organizations (or your clients') to better integrate experiences?
• Where do you believe the greatest opportunities for better integration still lie for you?
• How are you addressing the organizational and technical challenges required for better integration?
• How far down the path toward a more integrated, globally-optimized analytic perspective do you see yourself today?
• What's your favorite integrated experience story, for good -- or not so good?
• What resources have you found helpful for learning more / tracking what's going on in this area?
• What advice would you have for folks trying to push further down this path?

Suggestions for questions welcome -- just email me via the link at left.

----------

Postscript:  a recap of the panel on the MITX blog

----------

By this point, many of you will be familiar with some of the more interesting and exotic examples of "integrated cross-channel product experiences", such as the Nike+ product/ service/ community.  But the approach has gone mainstream too.  Here's a recent example I experienced:

I went with my family to the "99" Restaurant in Centerville, on Cape Cod in Massachusetts (the one on Route 28).  Lying on the table was a pad of these forms:

  I texted my information in, and 24 hours later this appeared in my inbox:

I clicked through:

Store to text to email to web to store, all nicely connected.  Cool!  Hope we're back before it expires.  Otherwise we'll have to sacrifice another family member's phone.  (Maybe do that anyway, and ask for separate checks... Hey, times are tough!) 

This program is run soup-to-nuts for the "99" by an external service called Fishbowl Marketing.  It's pretty good! I'm hoping to speak with them about experiences and results with it.

A few observations:

• Instead of texting "99", they might have asked for "CapeCod", or "Cville" to track the geographic location of the signup, allowing them to compare signups to store traffic across regions. (Fishbowl might need an SMS address unique to the "99" client for this, and that's more expensive.)
• I may not get back to the Cape soon.  When I clicked through from the email to the landing page from home, they could have recognized where the request was coming from and returned a link to a map of nearby locations under the coupon.  I'm betting most folks won't explicitly update their preferences, especially with the unremarkable landing page copy exhorting them to do so (good testing opportunity!).
• I sometimes participate in other offers like this.  They could offer me an opportunity to personalize my offer based on crossing my cell number against other databases that also might have it, to see what other information might be there that might have helped them -- and me (analogous to caller ID).
• Some sample of users could (may?) have been asked to participate in an exit survey about their experience, perhaps for additional benefits.
• maybe implement a "Share this coupon with three friends and get another $5-off when they sign up" opportunity? (I did get follow up email messages on holidays inviting me back, and suggesting I invite friends; but when I clicked through to landing pages for inviting friends, the forms asked me to tell them who I was, again, even though I hadn't cleared cookies in the interim.)
• Also, they might have added links to their social media presences like Twitter and Facebook (for feedback, or to offer other promotions), below the coupon. 
• Finally, and I'm sure this is on Fishbowl Marketing's agenda, they might consider signing up with a mobile location-based service provider, like Foursquare or Gowalla, either as part of the vendor branded experience or via a "white label" application developed off those vendors' APIs.  That way the "99" could vary loyalty rewards granted according to number of check-ins, and take advantage of the viral marketing advantages of these services (your friends get notified when you check in at the "99", so maybe they stop by too).   Of course you say, are "99" customers leading-edge tech adopters with the latest 3G smartphones, with OC(I)D (Obsessive Check-In Disorder)?  Not yet, but what were we saying about Facebook a couple of years ago?

What's your favorite example?  Hope to see you tomorrow morning!