Evan Schuman's StorefrontBacktalk

In the kind of move you’d expect from Walmart (NYSE:WMT) or Amazon (NASDAQ:AMZN) but not a retailer who couldn’t make the NRF’s Top 100 list, beauty products retailer Sephora has acquired a software startup. Last Wednesday (Aug. 7), the 1,300-store chain said it has acquired Scentsa, the company that developed Sephora’s touchscreen Fragrance Finder in 2008 and its Color IQ system last year—a pair of in-store kiosk systems that have become signature features for Sephora.

While it’s not surprising for a $2.6 billion retailer to develop in-store technology like that, no one seems to think of actually buying the startup as something most retailers can do. In part, that’s because most chains are deeply marinated in the buy-don’t-build concept, which makes perfect sense when it comes to ERP but no longer really holds for non-enterprise software. Besides, we’re not talking about acquiring IBM here: Sephora said the acquisition will add eight or ten people to its payroll.

That’s how small some startups are. What Sephora didn’t publicly disclose was how much it paid, but Sephora chief digital officer Julie Bornstein did have a short list of reasons for buying Scentsa, according to All Things D: To achieve long-run cost effectiveness, to develop new in-store products more rapidly and to keep the tech from winding up in the hands of competitors.

The first two are obvious and beside the point, because buying your software vendor—even a really small one—is the ultimate form of technology lock-in. Want to change platforms? You no longer have to just convince business-side users and managers. Now you have to explain to the CFO why this acquisition has suddenly become worthless.

So yes, you get to be first in line when you’re the only customer in line. And if you’re already a major customer, you won’t be paying to keep some investment fund fat and happy, so over time the purchase may break even.

But the real purpose is locking out the competition. When Sephora didn’t own Scentsa, the chain could implement the technology in its own way. But competitors could get access to the same underlying technology, and that could erode Sephora’s differentiation.

Yes, Sephora would still be Sephora, with all the high-touch customer experience and the brand value that the chain has built up. But at a basic level, there’s a real risk that it could all go away if someone else has access to the technology and enough money to try out-Sephora-ing Sephora.

Or at least there was that risk. Not anymore.

In fairness, none of this would likely be possible without a few other elements in the mix. For example, it probably helps that Sephora’s U.S. operations are based in San Francisco and Scentsa is just north of San Diego, both software startup hotbeds.

It’s not just that Sephora’s executives have rubbed elbows with startup types and understand the mindset. It’s also that they’ve seen how small a startup can actually be.

Remember, what used to be called a dog-and-pony show is for the purpose of making a vendor look much more impressive than it actually is. With a smart marketing chief and a budget that the company can just barely afford, a startup can appear big enough to convince much larger companies that it’s not the wing-and-a-prayer operation that’s one company-picnic-bus-accident away from nonexistence that it actually is.

And that, of course, is also a big reason most chains wouldn’t even consider buying a vendor. What, are you kidding? These guys are way outside the range of anything you could ever afford—right? Maybe not, after all.

Because once you get past the smoke and mirrors, the don’t-do-it-yourself commercial IT mindset and the fear of a very different way of thinking about IT, a retailer buying a startup comes down to something a lot like hiring a bunch of programmers for a skunkworks project. It’s just that calling it a “startup” gives that project a big head start—and a price tag that you won’t want to publicly disclose either.