Most companies try to avoid problems. Experian actually goes looking for them. In fact, it has set up a specific unit – Experian DataLabs — to actively seek out unresolved problems its customers are having and use them as a launchpad to seek out new opportunities and create new products. In doing so, it has been able to act more like a startup than a global data giant.
Conventional wisdom says that you need to run a big company differently than a startup and there’s a lot of truth to that. But for large enterprises seeking to grow by exploring new lines of business, thinking more like a startup makes a lot of sense.
Steve Blank, who pioneered the concept of the “lean startup,” has often written that “no business plan survives first contact with the customer.” That’s why he urges startups to “get out of the building” and talk to potential customers before beginning product development in earnest. There’s no use going to engineers with detailed product specifications before you really know what the customer wants.
That’s great advice and not just for startups. In fact, in talking to executives at Experian, it seemed clear to me that the techniques that Blank advocates, such as focusing on customer development before produce development, creating a minimum viable product and iterating and pivoting to a new business model, can actually be done more effectively in an established business.
Part skunkworks, part research lab, Experian DataLabs keeps a running list of the data problems customers want them to solve. As Eric Haller, Global Head at Experian DataLabs, told me, “We regularly sit down with our clients and try to figure out what’s causing them agita, because we know that solving problems is what opens up enormous business opportunities for us.”
Here’s a typical example. At a meeting with a large bank, one of the senior executives said, “You know we have a problem that’s really giving us trouble. We have a lot of newer businesses that come to us for credit and we need to do due diligence on them. So it’s an incredibly labor intensive process for us to verify whether they are a good credit risk.”
Haller’s team was able to put together a prototype and present it to the client within 90 days. It wasn’t perfect, the system could only identify 20% of the bank’s potential customers that had no discernable credit history, but that was enough to show the potential of their approach.
This is what is known in the lean startup world as a minimum viable product. Its purpose is not to wow anyone with exquisite design or top-notch performance, but rather to test a hypothesis. In this case, Haller and his team only needed to know whether if the customer would indeed be willing to pay for the product they had asked for.
Another tenet of Blank’s philosophy is that you initially build a product for the few (or in Experian’s case, the one), not the many. It is the passionate early adopters who help you to gain traction, see what works and what else may be needed to make the product successful. This isn’t market research, but hands-on problem solving.
And that’s exactly what the DataLabs unit is set up to do — solve problems. It doesn’t try to reimagine the future or parade nifty gadgets in front of the business press. Instead, it has dozens of PhD level data scientists working out of three separate labs on different continents that can supply impressive analytical horsepower to whatever might be giving a customer “agita.”
Once the DataLabs team validates the client’s interest—due to intellectual property considerations it almost always asks for a signed agreement before going beyond the minimum viable product stage—it begins to co-develop the product according to the client’s specifications. This tends to be an iterative process, with a number of versions going back and forth.
In the case of the credit product described above, over the next three months new features were added, such as a more helpful user interface, integration with other systems like auditing, and customized analytical options. Many of these improvements would not have been possible without the client’s input. Performance improved as a result; now the system was able to verify 50% of the “no hits” that were frustrating the bank.
At this stage, the team’s new product is presented to one or more “client advisory groups” inside Experian to see if there is more general interest. (Each group represents a functional area like credit, fraud, or marketing services.) The additional consultation from those groups also makes it possible for the team to pivot to different functionality, customer segments, or revenue models, if needed.
Once the product and the business model has been validated by Experian’s current customers and further input is taken from Experian’s client advisory groups and internal marketing staff, it is ready to be rolled out to the larger market. That’s when the rest of the Experian organization gets involved.
Engineers scale up the technology to ensure that it can work in a larger environment. Product managers work on issues such as pricing, legal compliance and positioning. Sales staff are trained so that they can handle client questions and a promotional campaign is designed and executed. Every aspect of the pilot project and the business model is refined and strengthened.
This process can take anywhere from three to twelve months, depending on how much integration needs to be done with Experian’s existing systems. That’s not particularly fast by the standards of a startup, but it’s not altogether slow either and it brings all the resources of a $4 billion company to bear, something that no startup can match.
In the case of the credit product, it will be formally rolled out this summer under the brand name BusinessVerify, about six months after the decision was made to move forward, which is roughly the average timeline for DataLabs projects. At the time of launch it will already have three paying clients and an army of salespeople with established relationships supporting it.
Clearly, large enterprises have things that slow them down. They must serve the present. Things are expected of them. They have to keep customers, employees and other stakeholders happy. But while small, agile firms can move fast, larger enterprises have the ability to move deliberately. They have loyal customers and an abundance of resources. Consider: startups usually only get to make one bet, but Experian DataLabs has a dozen or more disruptive projects going on a regular basis.
By assigning lean startup-like projects to a dedicated group of skilled people, big companies can get the best of both worlds, leverage their big-company assets to foster entrepreneurial agility.
Powered by Facebook Comments