Caliber: Closing the Loop between Online and Offline Purchases

A few months ago I did some consulting work for a startup in the Seattle area called Caliber Data. At the time they literally swore me to secrecy (with NDAs) about what they were doing. They brought me in to kick the tires of their model and give them some outside perspective on aspects of the local search market.

In a post in late March I alluded to them obliquely as a company that had developed a provocative system for closing the loop between online advertising and offline transactions. (In fact the post is so obscure that I can’t find it now!)

Regardless, these guys are now out in the world talking to potential partners and I’ve been told that people won’t show up at my house if I write about them now. 🙂

Mike Orren of Pegasus News guessed what it was when I did that original post. He inferred that there was a loyalty card at the center of the system. At the time I could neither confirm nor deny the guess but I told Mike he was a very smart guy.

Caliber’s system ties online advertising to the point-of-sale with a consumer loyalty card (but can also capture cash transactions). The merchant pays a marketing/success fee upon an actual consumer transaction. This is true CPA for the merchant – a sale rather than a click, even a call, which is a proxy for a potential sale.

The risk is transferred – on the continuum from CPM to CPA – from the merchant to the publisher or distributor. This is going to be scary to some publishers but it’s consistent with other moves in the local market (e.g., pay per booking). Google is testing CPA and Snap, Jellyfish and others have been using CPA models for some time. E-commerce affiliate models have existed since the beginning of the Internet.

I previously speculated about some benefits of CPA models for locally oriented publishers:

There are differences among all these models but they’re essentially seeking to deliver better prospects or actual sales with correspondingly higher ad/commission rates. What you can get for a click and what you can get for an actual customer are quite different.

(Citysearch EVP Scott Morrow spoke to me awhile ago about moving to a model that charges different rates based on “lead quality.” There’s a sales challenge there perhaps but it’s interesting to consider the continuum: clicks on the left and commission-based CPA on the right, with email, lead gen, chat and calls in between.)

Generally, there’s less of a sales challenge with CPA (unless its definition is variable) because the business understands it’ll pay only upon customer acquisition or its near equivalent. It also addresses click fraud, although I don’t think most SMBs think about click fraud that much. The issue is really “what is a click?”

Widespread adoption of a CPA model could potentially boost local online revenues and make the disparity between online and offline ad rates, if the publisher is also a traditional company, less than they are today with print vs. clicks for example. It also eases the pressure on publishers to have so much traffic volume to make money.

Back to Caliber.

When I first heard about what they were doing, the thing that really captured my imagination, beyond the CPA dimension, was the idea that you really could track online ads to the point of sale. Coupons, which have been around forever, also do this. But for reasons that are not entirely clear, the online coupon market is still immature.

There is something inevitable or inexorable about what Caliber is trying to do for small businesses. The challenge is bringing all the moving parts elegantly together – overcoming the “chicken and egg problem” that plagues many such new ideas. You need the distribution online to interest the merchants and you need merchant participation to get consumers interested, etc. This was a problem with local search itself in the beginning.

Caliber has told me it is currently in discussions with a number of companies – household names and lesser known players — to address these issues, including one major Credit Card issuer. Caliber’s folks say they’ve got the major pieces of the system (and related IP) in place.

I’ve argued relentlessly that the future of online shopping is offline: meaning that the dominant “going forward” model is the Internet driving in-store transactions. That’s always been true for service business and traditional yellow pages advertisers. (Given that fulfillment has to happen locally.) But it’s also true for retail. See, for example, this piece in the NY Times about e-commerce growth slowing. That doesn’t mean that the Internet’s influence is waning – exactly the opposite – it means online purchasing is not going to take over as once thought.

As the Internet’s influence grows but e-commerce slows, the central challenge for everyone in online advertising is how to track that online influence on offline buying. Aside from coupons and calls (to a lesser degree) there’s a “cliff,” after which the consumer becomes invisible and nobody really knows what happens in the local market where the transaction inevitably takes place. While there are convoluted ways to draw inferences and after-the-fact consumer surveys, there’s no real-time tracking beyond what I’ve mentioned.

Caliber’s system makes that entirely visible for both the publisher and the small business advertiser (maybe nationals too). It offers the potential to track the entire process from the search engine or directory or other online consumer destination site right to the local cash register itself.


11 Responses to “Caliber: Closing the Loop between Online and Offline Purchases”

  1. Mike Orren Says:

    Thanks for the shout, Greg.

    I like the model in concept, but I don’t believe it will ever work or scale in execution with a separate card.

    This sort of thing can be achieved with users’ existing credit cards. It’s not easy or cheap, but it beats special readers and getting another card into a users’ wallet.

    I’m also curious about how they plan to acquire users. I don’t see how you can do that without publisher/distributor participation, and this is a model very disruptive and frightening to traditional publishers — I know. I spent a couple years trying to sell it to investors/media cos.

  2. Greg Sterling Says:


    As I understand it it can be tied to existing credit cards or a phone number, like grocery store loyalty cards. So the physical card isn’t as important as participation/registration in the system.

  3. E-Commerce Slowing, Internet Influece Growing « Screenwerk Says:

    […] Screenwerk Greg Sterling’s Thoughts on Online and Offline Media « Caliber: Closing the Loop between Online and Offline Purchases […]

  4. Lee Wang Says:

    Thanks Greg for the post.
    We at Caliber are focused on the following –
    1. Serving business segments otherwise not using or getting benefits from today’s online advertising under current charge models with current complexities. It is about extending online advertising to places where it has not been working yet.
    2. Creating a service that also allows providers to succeed with us in gaining their footholds in merchants and consumer markets. It is all about connecting the dots rather than reinventing the wheels.
    Caliber Data

  5. Boris Mordkovich Says:

    I have 2 concerns with the overall CPA model:

    1) Isn’t CPA model just as susceptible to click fraud as PPC is? Fraudulent transactions, refunds, etc. – that could all play a role. Also, how exactly is CPA model different than all of the affiliate networks we have at this point?

    2) In order to scale something like this, you’d be faced with the problem of trusting that all of the advertisers put the code on their website necessary for conversion tracking. Is that really something feasible for 1000s and 1000s of advertisers?

    Boris Mordkovich

  6. links for 2007-06-20 « harbour 7 Says:

    […] Caliber: Closing the Loop between Online and Offline Purchases « Screenwerk (tags: ad_response metrics)   […]

  7. Jon Karlin Says:

    In response to Boris’s comment on CPA fraud:
    Unlike other CPA models, Caliber uses the offline transaction (swipe of a member’s transaction card at the point of sale (POS), to close a transaction. While we are not limited to a credit or debit card based system – (Caliber can work with rechargeable stored value, a.k.a. “Gift Cards”) or even a unique id that serves as an alias, there is probably not other area of technology where anti-fraud has been so heavily engineered and tested. Our connection to the advertiser is through a code that is generated by affiliate advertisers. We use this to “open” a transaction which is later closed after a sucessful offline purchase (card swipe at POS). We have very strong interest from the major search engines in this model – even though it requires a click or hover (“action”) from a user because we present them with a way to penetrate the relatively untapped local offline commerce opportunity. This represents one of the biggest opportunities for search engine/advertisers since unsubscribed local ad inventory is running VERY high. We provide incentive for each member in the value chain, from the advertiser, to the transaction processor, the merchant and the consumer. Only when there is a valid transaction and payment has been tendered do we credit each of these parties (except for the merchant who benefits from the sale and from our online ad placement service). While we don’t claim to eliminate completely any potential for fraud, we create incentive for each member to behave in a way that supports our system (for example the consumer will insist on swiping the card since they get redeemable points). There is a very interesting and relevant article on CPA and fraud (which is ironically titled in a seemingly adverse way) at the following link. Net, net – we believe that by connecting the offline transaction processing world with the online ad world – we have one of the best solutions to fraud and THE best solution for penetrating local online advertising. see:

  8. Per Wassgren Says:

    As a comment to Boris comment about possible fraud within using a CPA-model, I was running a non-transaction based CPA campaign with Zanox in the US this summer when I discovered som strange sign-up patterns and after having alerted Zanox about the problem, they investigated and found that 75% of the traffic was actually fraudulent where all hits came from (probably) related IP-addresses in China. My guess is that it was a number of humans behind the traffic rather than some software but I don’t know. Zanox denies that anything could have been done to prevent the fraud but looking at the IP and sign-up data which occured in fairly symmetrical patterns, I disagree.

  9. Offline Conversion Tracking « Screenwerk Says:

    […] reporting or proxy behavior (clicking on maps, send to phone, calls). Some folks (i.e., AOL and Caliber data) are working on loyalty card POS tie ins that will enable tracking down to the […]

  10. Solving the Online-Offline Problem « Screenwerk Says:

    […] the Online-Offline Problem Caliber Data is a Seattle-area startup, run by former Microsoft employees, which has been seeking to bring […]

  11. Martin Kahn Says:

    Jon, I ve been reading about your company and i think you are on the way.

    I have two concerns. 1. As I understand Caliber tracks offline transactions using existing credit or loyalty card, but, are you tracking transactions at SKU/product level? That would be really interesting.
    2. I didnt understand when a consumer signs in or register his cards. Is when they click on the advertisign? How do you identify a consumer when he clicks on an advertising?

    Congratulations, I believe your model is the future of Online/Offline Marketing.

Comments are closed.

%d bloggers like this: