An innovation lab is a small team, usually with its own budget and often its own building, that a company sets apart from its day-to-day business to invent, build and test new products, services or business models. It runs on startup rules: short cycles, cheap experiments and permission to kill ideas. In the end a corporate innovation lab is judged on one thing, which is what it puts in front of paying customers.
That last part is where most of them stall. In 500+ launches since 2016, Blazon's corporate work has kept landing at the moment a lab product has to leave the building, including HelloEar by Xiaomi, Absolut Art's Hangsmart and Pillo Health, later acquired by Stanley Black & Decker.
The lab is rarely the problem. The prototype usually works. What is missing is a price, a buyer and a launch.
What is an innovation lab?
Search the term and you get three different things. Harvard Innovation Labs, which has supported student and alumni ventures since 2011, is a university lab. City and coworking "innovation labs" are shared workspaces for local founders.
Both are useful. Neither is what a company means when it says it is building one.
A corporate innovation lab is a unit inside an established company whose job is new revenue the core business would not create on its own. It might sit in a separate building, report to a chief innovation officer, or be ten people borrowed from three departments.
Three questions matter more than the room: what the lab is for, who owns its output, and how that output reaches a customer. The third decides whether the lab is a business or a showroom.
The main types of corporate innovation lab
Most corporate labs fit one of five models. They differ in time horizon, in who pays, and in what "shipped" actually means. Picking the wrong model is the first way a lab fails, because each one needs a different route to market.
| Lab model | What it does | Example | What shipping means |
|---|---|---|---|
| Research lab | Long-horizon science the core business cannot fund on a quarterly clock | X, Alphabet's moonshot factory | A project graduates into a standalone company or a Google technology |
| Product lab | Builds new devices and products for the parent to sell | Amazon Lab126 | A product on sale under the parent brand |
| Venture studio | Creates new brands in categories the parent has never sold in | P&G Ventures | A new brand earning real retail sales |
| Operations lab | Prototypes changes to the parent's own stores, service or packaging | Starbucks Tryer Center | A change rolled out across the parent's own locations |
| Open innovation lab | Scouts and pilots startup technology instead of building in-house | Corporate accelerators and venture-client programs | A startup's product bought and deployed by a business unit |
The operations lab ships most reliably, because its customer is the parent company. The venture studio and the product lab have the hardest job.
They need customers who have never heard of the product. That is a launch problem, and more engineering does not solve it.
Innovation lab examples that hold up
Most "top innovation labs" lists are directories of logos and founding years. These examples come from the companies' own pages and announcements, and each teaches something about getting to market.
X: graduation is not a market. X, the moonshot factory, publishes its record openly, with current, graduated and discontinued projects side by side. Two of its 2018 graduates show the spread.
Wing started in 2012, made its first real-world deliveries in rural Queensland, Australia in 2014, and graduated as an independent Alphabet business in July 2018. X says the team has since conducted more than 500,000 commercial flights.
Loon graduated the same year. By 2021 it was closed, and X's own project page gives the reason plainly: "the path to commercial viability proved much longer and riskier than hoped."
Same lab, same graduation year. One found a commercial delivery business. The other ran out of road before it found one.
X lists Glass as discontinued. Its Glass page says the team's "living lab" approach was expanded to the public in April 2013 through the Glass Explorer Program. The lesson for any lab: once a product is in public hands, the market treats it as a launch, whatever the internal label says.
Amazon Lab126: launch to a small room first. Amazon introduced Echo in November 2014 and, by its own account, offered it "on an invitation basis" to customers who wanted to help shape the device. Over the next seven months Alexa added dozens of new features. Only then, in June 2015, did Amazon open Echo to all US customers at $179.99, shipping July 14.
That is a launch designed as a test. The invitation list limited the damage if the product disappointed, and it fed the team real-world use before the full launch.
Most corporate labs do the opposite. They test privately for years, then launch to everyone on one day.
P&G Ventures: kill most of it, back the winner. Procter & Gamble launched P&G Ventures in 2015 as an internal startup studio aimed at categories P&G had never competed in. Its first major success was Zevo, the bug-control brand launched in 2017.
P&G says two decades had passed since it last introduced a new-to-the-world brand that reached $100 million in sales, which was Febreze in 1998. Zevo reached that milestone again.
The less-quoted part of the story is the pruning. P&G says the unit began as six organizational units and 29 brands, and is now one team focused on a select group of brands. A lab that never kills anything is not testing anything.
Starbucks Tryer Center: the parent is the customer. In April 2024 Starbucks put new cold cups into stores across the US and Canada after testing them in a laboratory inside the Tryer Center, its innovation hub in Seattle. The cups use 10 to 20 percent less plastic. Baristas rated each iteration in test and store settings, customers answered surveys through QR codes, and Starbucks says the new cups cost less to make.
That is the easiest kind of lab win, because the buyer, the channel and the rollout already existed. A lab that makes something the parent's own stores need has its route to market on day one. A lab that makes something new to the world has to build that route itself.
Why most corporate innovation labs fail to ship
Pages ranking for this topic quote figures like "up to 90 percent" of innovation labs failing. They come without a sample, a method or a definition of failure, so they are not in this article.
The documented numbers are more useful. Capgemini surveyed 1,700 people at 340 billion-dollar companies across eight countries in 2017. It found that 87 percent had "a lab or space dedicated to innovation", but only 17 percent said innovation was carried out across the organization, beyond the innovation center.
The same report says many organizations "lack established processes for managing an idea through to commercialization." That line is the whole problem. Across 500+ launches, Blazon sees five patterns behind it.
The lab is measured on activity. Ideas generated, hackathons run, prototypes demoed. None of those numbers involves a customer, so a lab can hit every target and still ship nothing. This is innovation theater, and it survives exactly as long as the budget does.
Nobody owns the launch. The lab builds the product, then hands it to a business unit with its own targets and bigger brands. The handover is where products go to wait. Blazon's analysis of why corporate innovation stalls at launch covers the closures behind this pattern, from Walmart to General Mills.
The launch is borrowed. Brand, PR, paid media and commerce all belong to other departments. Each one agrees in principle and schedules the work for next quarter.
The prototype meets the parent's quality bar. A first version carries the corporate logo, so legal, brand and quality teams hold it to the standard of products that have shipped for decades. A startup with the same idea launches the rough version and learns from buyers instead.
Nobody ever asked for money. This is the big one. Labs test whether people like a product, understand it or would consider it, and rarely whether they will pay.
Blazon calls the fix The Deposit Test: a buyer who will not put down a small deposit now will not pay full price on launch day. It is also the first thing Blazon's corporate innovation team runs on a lab product.
The innovation workshop: where most lab work starts
An innovation workshop is a structured, time-boxed session where a small cross-functional team takes one defined problem and works it through to ideas, a prototype and a decision. Most corporate lab projects start this way. Most workshops also end too early.
The best-known format is the design sprint. GV describes it as a five-day process for answering critical business questions through design, prototyping and testing ideas with customers, and says Jake Knapp began running sprints at Google in 2010.
Monday maps the problem. Tuesday sketches competing answers. Wednesday picks one, Thursday builds a realistic prototype, and Friday puts it in front of real customers.
Hackathons are the looser version. Microsoft held its first company-wide hackathon in 2014, and one project from that culture, Skype Interviews, went from hackathon to shipment in one month.
Seeing AI, Microsoft's app that can read menus and documents, launched in 2017 after the blind engineer behind it was given two months off his day job to work on the presentation that became the product. Each of these had time, a sponsor and a route to ship. The workshop was only the first day.
A sprint tests whether customers understand a prototype. It does not test whether they will pay for it.
Blazon's version of the workshop ends with three outputs: a price, a draft launch page carrying the headline that will run in the ads, and a date for a paid demand test. A team that cannot write an ad headline a stranger would click has not defined the product yet, and no amount of extra engineering fixes that.
How to set up an innovation lab that ships
Most guides answer this with org charts and floor plans. The decisions that matter are fewer and less glamorous.
Pick the model before the people. A research lab needs scientists and patience. A venture studio needs brand builders and a launch budget. Hiring for the wrong mandate burns the first year.
Name a commercial owner. One person, with a revenue or adoption number against their name, from the first week. Without that person, the product has nobody to argue for it when budgets are cut.
Fund the launch inside the lab. A request into a business unit's marketing plan competes with that unit's own brands. A budget held by the lab gets spent on the launch.
Set the evidence threshold up front. Decide what result moves a product to the next stage: a number of paid deposits, a pre-order rate, a pilot customer signing. A good reaction in the steering meeting is not a threshold.
Choose the route to market at kickoff. The parent's retail channel, a direct-to-consumer store, a pre-order or deposit campaign, a licensing deal or a spin-out each imply different products, packaging and timelines. Blazon's go-to-market strategy guide covers the sequencing.
How to get innovation lab products to market
Getting a lab product to market is a launch with corporate governance attached. Three cases from Blazon's own work show the moves that matter most.
Borrow the parent's credibility without disappearing behind it. HelloEar made custom-fit earphones with Xiaomi behind them, and it was walking into an audio category owned by established brands. Lean on the Xiaomi name too hard and the product disappears behind the logo.
Blazon reduced the fit technology to a single benefit a buyer could repeat, earphones shaped to your own ears, and put paid media, press and influencer outreach behind it. HelloEar by Xiaomi went on to $1.2M in revenue at launch.
For a corporate team, a launch like that is a public, priced test with a result nobody inside the company can argue with.
Explain the new category in plain words. Pillo Health, later acquired by Stanley Black & Decker, built a home-health robot that helps people manage their medication. Nobody had a mental slot for that product, so its story had to answer three questions quickly: what is it, where does it fit in my day, and why should I care.
Blazon wrote that story and ran the PR so the robot came across as friendly rather than clinical. It went on to 15K+ units pre-ordered, 120+ press features and $2.5M in sales.
Lab teams find this step hardest because they know too much. They describe the technology, and buyers want to hear what changes in their day.
Run the launch as one program. Hangsmart is a small device that makes hanging art simple, and it came from Absolut Art, the spirits company's direct-to-consumer art platform. Blazon put the strategy, campaign creative, platform launch and press, the last run jointly with RVD Communications, on a single plan with one message and one place to buy.
The Hangsmart launch picked up 100+ press mentions across outlets including HYPEBEAST, WIRED and Digital Trends. In most corporate labs, those workstreams would have been four separate requests on four separate schedules.
Sell it somewhere it is the main event. A lab product sitting beside the parent's bestsellers gets compared with them almost immediately, on price, volume and margin. A store of its own buys the product time, hands the team its own customer data and lets the lab set the price.
Prove demand before the big capital request. A deposit page with paid traffic costs a fraction of a tooling order and returns a number a board can act on. Blazon's demand validation ladder walks through each step from prototype to pre-orders.
What an outside launch team adds
Blazon is the corporate innovation agency that takes lab products to market. It works alongside an innovation lab, a venture studio or an intrapreneur team and takes on the commercial half: pre-order and deposit tests, brand and positioning, go-to-market, the launch itself, a Shopify store in the company's name, PR and paid media.
It is the same playbook behind Blazon's product launch agency work for founders, scoped to brand sign-off, legal review and procurement.
An outside team cannot fix governance. If the lab has no mandate to sell and no budget of its own, the sponsor has to settle that first.
What it does fix is speed and evidence. The people who write the positioning also build the store, buy the media and pitch the press, so the plan meets real buyers within weeks. For a product stuck between the lab and the market, start with Blazon's corporate innovation launch team.
Most labs are built to answer whether something can be made. The ones that last are built to answer whether someone will buy it, and they ask that before the prototype is finished.
Frequently asked questions
What does an innovation lab do?
An innovation lab finds, builds and tests new products, services or business models away from the company's day-to-day operations, using small teams, short cycles and cheap experiments. A well-run lab also owns its route to market, so its work reaches paying customers instead of stopping at a demo day.
What is a corporate innovation lab?
A corporate innovation lab is a semi-independent unit inside an established company that creates new revenue the core business would not produce on its own. The common models are research labs, product labs, venture studios, operations labs and open innovation programs, and each needs a different route to market.
What are some famous innovation labs?
Well-known corporate examples include X, Alphabet's moonshot factory, which graduated Waymo and Wing. Amazon Lab126 is the team behind Echo, P&G Ventures created the Zevo brand, and Starbucks runs the Tryer Center in Seattle. On the university side, Harvard Innovation Labs has supported student and alumni ventures since 2011.
Why do innovation labs fail?
Most fail at the handover rather than the invention. Capgemini found that 87 percent of large companies had an innovation lab or center, but only 17 percent said innovation happened beyond it. The usual causes are measuring activity instead of buyers, no named owner for the launch, a launch budget held by another department, and no early test of whether people will pay.
What is an innovation workshop?
An innovation workshop is a structured, time-boxed session where a cross-functional team works one defined problem through to ideas, a prototype and a decision. The best-known format is GV's five-day design sprint, which ends by testing a prototype with real customers. A workshop that ends without a price and a test date usually ends in a slide deck.
Is an innovation hub the same as an innovation lab?
The terms overlap. An innovation lab usually builds and tests products for one parent company, while an innovation hub is more often a shared place, run by a city, university or coworking operator, where startups, investors and corporate teams meet. Some companies call their own lab a hub, so judge it by what it produces.
How do you get an innovation lab product to market?
Put a price and a deposit or pre-order page in front of real buyers before the big capital request. Then launch it as one program, with positioning, a direct storefront, PR and paid media on the same plan, run by a commercial owner with a budget. Decide in advance what result moves it into a business unit, a spin-out or a shutdown.
