Back to Gartner for Marketers Blog

What Oracle’s Acquisition of Maxymiser Really Means

By Martin Kihn | September 22, 2015 | 3 Comments

It’s not what you think.

Now that the dust has settled on one of the quietest marketing technology acquisitions of the year – that of Maxymiser by Oracle on August 20th – it’s time to pause and ponder what it means. Because Oracle has gone from zero to sixty-five in the mad-tech arms race more rapidly than anyone anticipated. (Mad-tech is my own combination of the phrases “marketing technology” and “advertising technology.” You’re welcome.) So it has got to tell us something.

And it does.

First, it tells us that mainstream journalists covering our space still do not know much about it. The Wall Street Journal headline was: ”Oracle to Acquire Mobile-Marketing Startup Maxymiser.” Now, whatever Maxymiser is – and it is many things, which we’ll get to – it is neither particularly mobile nor, in fact, a startup. (It was founded in 2006, and its charming British founder Mark Simpson told me it was profitable last time I spoke to him.)

And the redoubtable Business Insider launched a squib called “Why Oracle just bought a marketing company called Maxymiser” that included only eleven words on, well, why Oracle just bought a marketing company called Maxymiser. They were: “Chief Marketing Officers now command huge budgets for tech and software.” So there.

Second, it tells us that customer experience has won. What do I mean?

Let me set up my reveal by laying out 9 Business Insider-sized factoids about Maxymiser:

  1. It was founded by Brits and based largely in Europe until last year, when it invaded New York City as a beachhead in its westward march
  2. You can tell is was founded by British people by looking at the way they spelt “Maxymiser”
  3. It emerged a decade ago as a fiery competitor to Offermatica and Optimost, both of which were acquired by big companies with big marketing ambitions (Adobe and HP, respectively)
  4. So it is firmly in the A/B and multivariate (MVT) testing space
  5. Maxymiser’s main competitors in the larger leagues are Adobe (via the Target product), Salesforce (Web Predictions), IBM (Digital Analytics and Content Recommendations), HP Autonomy (Optimost) and Webtrends (Optimize)
  6. Among smaller players, it is flanked by Optimizely and Monetate, which offer SaaS-based testing and site personalization that is also easy to use
  7. In particular, Optimizely changed the game for testing and optimization by having clear pricing, point-and-click tweaks, easy integration with Google Analytics, permissioning, a free trial, actual self-service useability, and so on
  8. Unlike Optimizely, Maxymiser is definitely focused selling to the enterprise, with 300-odd clients clustering around those areas most interested in improving the conversion rates of things they sell on their websites – that is, retail, travel, finance
  9. Like many enterprise-aimed software products, it comes layered with high-touch professional services which the company claims are an optional nice-to-have and which rumor claims are bit more than that

And for symmetry, one factoid about the deal itself:

  1. Maxymiser gets Oracle’s cloud onto the website — it lets Oracle, at last, touch those magic moments when an individual goes from ad prospect (targeted via ad tech) to customer (targeted via mar tech); it bridges the mad-tech gap

Now, the big reveal. Remember my primary point: Maxymiser was a proud and mighty A/B MVT testing product. Got it?

Let’s examine the press release, issued by Oracle last month. Let’s look at it in the form of a word cloud:

personalization customer experience wordle

Notice anything barmy? Unexpected? Conspicuously inconspicuous? No?

Here’s something: Where are AB and MVT or any other type of testing? Nowhere, that’s where. They’ve been entirely and utterly subsumed by the master meme of customer experience and that, my marketing friends, is exactly what I mean when I say that CUSTOMER EXPERIENCE HAS WON.

So what? Well, for one thing, at least, three formerly respectable, separate marketing tech categories are now officially one. I refer, in order of age, to: Landing Page Optimization, AB Testing, and Personalization. These are all now the same category.

But the currents run deeper than that – so much deeper. Customer experience has become the dominant metaphor for marketing technology as a whole. It is the north star of the mad-tech universe. It is the organizing principle around which marketing- and ad-tech – mad-tech – have converged. This is neither good nor bad, and may be more good than bad. But it’s here. Oracle-Maxymiser proves it.

The Evolution of Personalization

Were I optimizing blog views, I’d have stopped there. But I’m not. I’m not an optimization platform, friends, I’m a blog-reader experience platform – and so, I continue. Let’s indulge in a bit of history mining, consider the morphology of personalization, touch on predictive analytics, and land our ski-doo somewhere in the near future of, say, 2016. Hip?

AB testing has grown up with mar-tech and ad-tech and mirrors its philogony. Its first incarnation tested versions of, say, landing pages against random samples of a single population and declared a winner. This procedure is still used. It’s why you see so many orange “BUY NOW” buttons all over the place – orange buttons won a lot of tests.

Of course, the mass market approach works on average, but it has an obvious flaw. Different groups of people exist within the mass market – let’s call them segments – and they may respond better to different messages. That’s why they’re segments. To press our point, it’s possible there may be a different “BUY NOW” button color that works better on one segment or another. Less trivially, there may be a different product page that works for different segments.

Maxymiser had a nice way to target tests against segments. And how are the segments themselves defined? A marketing analyst could define them manually, if the visitors were logged in or previously identified. Or they could be identified by what they have done during a session – that is, I browse around the Bernese mountain dog-themed tea cozies (to take a recent personal example) and I’m put into a group of proud “Bernese mountain dog fanatics.” This is called behavioral targeting.

The next peak for testing platforms to climb was the problem of scale. There are dozens, hundreds, thousands of things that could be tested on a modern site; and, since we can do segment-based testing now, what about all the segments? Data points themselves that can be used in targeting are also daunting: location (using IP address or even, at times, app location), time, behaviors, referring sites and traffic sources, info from the CRM system, ya ya yo.

This all calls for automation. Last year, Maxymiser launched a feature it called MaxPredict that did something called segment discovery, using unsupervised techniques to identify groups of site visitors, run tests, declare winners against different segments, and allow more rapid improvements. Auto segment discovery is a fiendishly difficult task, even today, and even test pilots like the 150 engineers at Google Analytics are trying to figure out how to productize it.

A trickier problem – which gets us right up to the living present in the personalization space – is cross-channel profiles and cross-device identification. The goal is clear enough: to be able to resolve identities no matter what device or browser people happen to be on at the moment, and to be able to run tests against individual on different channels. And so on.

So it’s easy to see now how “personalization” and “testing” turned into the same thing. Personalization engines – examples include RichRelevance, Certona, Baynote – use segments and behaviors to suggest content that is more likely to sell (or be read, or whatever), and they learn from mistakes. Testing platforms do exactly the same thing. They’ve converged.

The final step in our wonky walk here is from personalization to . . . you guess it: customer experience. Customer experience is the new personalization. In moving its mojo from humble “AB/MVT tool” to haughty “Customer Experience engine,” Maxymiser reflects exactly the industry’s shift from talking to itself about technology to talking to outsiders about love and humanity.

In other words, mad-tech is now like every other consumer-facing industry in the world. Ultimately, the marketers have won.

Comments are closed


  • Pat LaPointe says:

    Martin –

    All good. Nice summary. The technologies have begun to converge to enable “personification” (as I believe you Gartner folks call it). Segments get more micro all the time (don’t forget the psychographics and motivational drivers on the list of segmenting dimensions). But here’s where the next obstacle arises, as expressed by a Fortune 100 marketer recently: “sounds good, but my agency will charge me a fortune to develop all those variable creative executions necessary to really deliver a relevant customer experience, and I don’t have the budget or bandwidth to manage that circus.”

    The one piece of mad-tech missing from the convergence picture so far is the efficient and scaleable adaptation of actual creative messaging to match the segment. If the segments are all behaviorally defined in terms of predictive product/pricing needs, then the creative doesn’t matter so much. It’s a race to the bottom fueled by pricing comparison bots. But for anyone really trying to wean off “buy now and save” promotional pricing and make a more multi-dimensional brand connection, the message needs to resonate on some other levels that only words and imagery can convey.

    Fair to say that we still have lots of opportunity curve to climb between here and that asymptote. But a former mentor of mine once taught me something that seems really applicable here: technology enables, but imagination wins.


    • Akin Arikan says:

      Pat is raising a very practical and thoughtful point as always. The one exception I would say are product and content recommendations that are created automatically from website content however. That way, no creative team and budget are needed to craft the personalized experiences. Martech companies such as Boomtrain for example help publishers and retailers create emails that look handcrafted but are populated by an algorithm. This can be either a sideline within an email or the whole thing (with a pinterest-like design). In the background there is an algorithm that does segment discovery.

      Note; I am not affiliated with Boomtrain.

  • Great article and particularly struck a cord with me, an Oracle alum. I’ve heard the same challenge outlined by @measureman related to multiplying cost of creatives / content. I remember reading a recent article on Digiday that creative is one area of marketing / advertising that’s relatively untouched by “madtech”. That said, Persado provides a reasonably scaleable solution for auto ad copy generation.