Shoeboxed, the very successful Durham receipt-data-scanning company, made some noise about 10 days ago when they launched a new product enhancement, Price Deputy.

Price Deputy, now in beta, will alert you when someone in the massive Shoeboxed database of tens-of-millions of receipts bought a product for less than you paid for it. It will also summarize the retailer’s return and price-matching policies in hopes of getting you a better deal in retro.

The product will eventually expand and do more and be better, but even as it stands it’s a pretty cool idea. So cool, in fact, that I immediately became jealous.

Which is exactly what Tobi Walter, COO and co-founder of Shoeboxed, wanted. Not really. But when he and I finally sat down to catch up, I told him as such, which got a laugh.

In my mind, the exciting news here is that Price Deputy came out of a new initiative called Shoeboxed Labs, an experimental arm of Shoeboxed powered by the by-product of their core service, data from these tens-of-millions of receipts.

Every startup, no matter what stage or size, should be doing this kind of thing.

“We’re finally getting to a point where we spend 80% of our time on the core Shoeboxed product,” Tobi said. “We also want to spend 20% of our time on mindshare and what else is out there. What else can we do? Labs is everything that’s non-core. Core stuff gets worked on as usual – but this is stuff that doesn’t necessarily have to become core.”

Some features are essential. Making the product more accurate, easier to use, all those things that enhance the quality and the user experience, those are core features. Labs is about new directions.

“From a use case, it’s probably separate from our usual Shoeboxed experience,” he said. “Experimental stuff doesn’t always work 100% of the time, and the core product has to. Labs helps us test and build the product. Plus, developers are able to have a sandbox and not have to wire new features in to work with everything all at once, like paper documents, API. They can become standalone.”

It’s an essential part of any startup. Once the core product is launched, the work can’t stop there – the innovation can’t stop there. An entrepreneur will want to keep pushing the boundaries of the product and the company.

“You don’t know how big things are unless you try them,” Tobi said. “Price Deputy may someday become bigger than Shoeboxed, and we won’t know that unless we try. We don’t want to be outrun by any innovation in adjacent spaces. Maybe in five years there aren’t paper receipts anymore.”

But it’s not just about new products.

“We may find enough people want to pay for these services as part of core, and we add it on to the core product,” he said.

Several years ago, one of the first things I did when Automated Insights pushed out our first set of automated content was set up a directory called StatSheetLabs (back when the company was still known as StatSheet and focused primarily on sports). The contents of this directory would include three things:

1) Potential product enhancements we could muck around with.

2) Potential new product ideas we could scratch out and test.

3) Shit we thought would be fun to do.

StatSheetLabs would eventually become the launch pad for everything the company produced, and it still acts as that kind of primordial ooze today. In fact, there are about fifty things residing in that directory right now that I can’t even talk about.

Our labs has paid off many times over. In fact, part of my job is to take what we learn in labs and figure out how we can fundamentally change our product, our company, and the science of what we do.

Yeah, I’ve got a pretty freaking cool job.

We’re not the only startups doing this, obviously, but I’m always surprised at how few startups take advantage of the residue that gets left from their products or services, especially when that startup specializes in big data.

Google was a search engine. Amazon sold CDs and DVDs.

In fact, for all intents and purposes, ExitEvent is an informal labs product of the last startup I founded. The two are tangentially related at best, but until I started executing on the idea, ExitEvent would have remained a cool thought that would have been fun to try.

So let that be the lesson. If you’re a startup, even if you’re just one person working on one product, have a labs concept going, and spend time on it.

You never know what you’ll come up with.