Forrester Analysts Focus on Adaptive Intelligence DF14

It's no secret in the age of customer-centricity that marketers need to leverage data to enhance service levels. So how's that going?

Not that well, it seems. Two principal analysts for Forrester Research shared their insights today at Dreamforce in San Francisco.

"Every client we deal with has operational systems that just can't keep up," said John R. Rymer. "And they have to reinvest in those systems so they can actually act on the insights they get."

Friendly Advice

Rymer and his colleague, James Staten, had plenty of advice for business managers facing today's fast moving changes in the marketing arena. Most of it stemmed from the use of "adaptive technology," which Forrester defines as the real time, multi-directional sharing of data that delivers contextually appropriate, authoritative knowledge to maximize customer value.

Rymer offered three pieces of advice:

1. Innovate at Customer Speed. "Customers change their minds and change their tastes very quickly. If you're in B2B, markets move. If you're in the government, citizen patterns move."

2. Know Your Customers Better than Anyone Else. "Whether they're students, citizens, consumers, know them really, really well so you can keep up."

3. Turn Your Data Into Money. "Everybody's got valuable data."

Of course, that's all easier said than done, but the analysts had plenty of guidance based on what they see while consulting with client companies.

A key objective is to abandon the old ways of IT development. "We've got to deliver so quickly now that it really puts pressure on the old way of doing stuff. It just doesn't work very well," Rymer said.

The good news is that building apps based on third-party APIs not only saves time now, but allows for rapid revisions as customers react in the marketplace.

"Part of the value of APIs is you can move faster because you just reuse things that already exist, that somebody else has already built," said Rymer. "And you don't have to build it and maintain it yourself."

The apps built that way are "composites" of your own code and the APIs. "Each composite can evolve on its own track so it's much easier to change these applications rapidly over time," he added.

One key, Staten said, is in "really engaging your customer in any way you can, and using technology at the time and moment necessary to give them the value that is relevant to them at that point."

Testing Results

That requires fast A/B testing of marketing actions to assure they are working. "If you're not already doing A/B testing today, you need to do more and more of this, getting capabilities out there and letting people tell you what works and doesn't, so you can move really quickly," Staten said.

"As soon as you have created an experience for using a piece in context and you've taken an action, start monitoring that action," he said. "Was that the right action? Did it drive the behavior you're looking for? Did it not? Then change it, and move really quickly."

This means that in addition to have marketing automation technology to help create a timely communication with customers, you also need "responsive" systems to gauge their response. Relatively few companies have such systems in place, the analysts said.

Use Your Data

Most companies have data on customers, but the analysts pointed out the importance of blending that with third-party data to get clarity into the customer's mindset. Staten offered the example of Starbucks, which found the sales of pastry were falling in its New York City stores.

The Seattle-based coffee chain bought GPS-generated location data generated by AT&T and a startup that helped to track customers after they left the store. It found that the customers were making a beeline for McDonalds, preferring a hot Egg McMuffin to Starbucks' cold baked goods. Starbucks responded with its own breakfast sandwiches.

"There's a ton of data out there to help understand your customers. There's a whole world out there, starting with location," Rymer said. "The challenge in an exercise in real-time systems, in real-time information processing. The cloud is essential to manage this scale. But you're also dealing with information that's moving in time. And for a lot of us, that's new."

Don't Be Creepy

In the question-and-answer session after their talk, audience members asked how far companies can go in using customer data without it seeming creepy.

"It's going to vary," said Rymer. "Part of why you want to get involved with contextual customer interactions sooner rather than later is because there's no formula. You're going to discover it through experimentation and feedback from your customers."