Select Page

“Startups that succeed are those that iterate enough times before running out of resources.”  Eric Ries

In other words, how do we learn fast to build evidence of value we need in order to convince others to give us the next level of resources? Startups do it every day because they have no choice. They do what it takes to learn quickly or they die. In big companies we have other choices and higher order demands that require us to do things ‘properly’. These behaviours can materially throttle a company’s learning velocity.


Here are two perfectly sensible things that happen in companies every day. They are both caused by talented people doing their job brilliantly but accidentally crushing innovation and learning velocity.

    • Situation: A customer development team wants to run a channel experiment to find optimal fit with customers, value proposition and channels. They create a website that reflects their first hypothesis and explore what it takes to get the experiment live.
    • Anti-Pattern: The marketing team has the epic task of pulling together the communication activities of thousands of employees into a consistent message and brand. A discussion with this busy team now commences. Naturally, they prioritise this insignificant project alongside a multi-million dollar campaign that will launch soon. Communication is slow and not across the purpose of the experiment.
    • Outcome: Many weeks later (if the experiment actually makes it into the world), a website is released. The insights from the experiment are diluted because messaging and visuals were tuned away from the fit that the team wanted to test towards the static brand of the main company. What a startup would run as a full cycle in 48 hours is now only just able to begin gathering evidence.
    • Situation: A customer development team wants to run a problem-solution fit experiment for a brand new product and wants to send an email to 10,000 customers to measure their response.
    • Anti-Pattern: The legal team protects the company from making decisions that may open it up to attack from the external world. The team “checks with legal” and receives a series of questions back.  A number of potential risks are surfaced and the team is asked to resolve these before proceeding. A long email thread opens up, the email copy is adjusted to lower any risks in the offer, making it very difficult to measure whether customers valued it or not. The team drifts away from the core an starts to feel a bit worried that their work could get the company into trouble.
    • Outcome: The new idea is killed by the team as they loe sight of the value and don’t want to take on the risk.

The solution of course is not to let the company run wild with the brand and open up risk at every opportunity. The solution is to design a process around early stage innovation that manages these streams effectively. Facebook does it really well.

“There are tens of thousands of versions of Facebook running at any one time. Facebook has a testing framework that lets engineers to run experiments quickly. At any given point in time, there’s not just one version of Facebook running in the world. There’s probably tens of thousands of versions running because engineers here have the power to try out an idea and ship it to maybe 10,000 people or 100,000 people. And then they get a readout on how that version performed compared to the baseline version of Facebook.” Mark Zuckerberg



From 2006 to 2011, Google ran a program called Google Labs. When we used products in Labs as a customer, we knew when we signed up that they may break, that they will certainly change a lot and may be discontinued altogether with little notice. We knew we were lab rats. Products like Gmail and Calendar came out of Google Labs and are now massive products. Other products like Google Wave didn’t make it. While Google Labs has now been closed, there are still options for customers to enable and disable access to beta software. A beta program or lab can create a frame around early stage innovation that big companies can manage risk within:

  • Reputation Risk: customers know that it may break or be discontinued. They don’t look at this product in the same way as more established products that they depend upon. Even if they grow to depend upon it, they knew the deal coming in.
  • Legal Risk: any product in a Labs program can have clearly predefined legal terms that give teams more leash and lower service levels than mainstream products. Labs programs can also allocate lawyers on the team that are excited about making the new stuff possible and dedicate their time to staying on top of how innovation firms are managing risk through new legal structures.

The substantial upside of a lab/beta program is that your firm will acquire over time a passionate group of early adopters that will be excited about trying the new stuff. One of the blockers we see frequently in large enterprise is a struggle knowing who to reach out to in early tests.

You don’t need to be a digital company to have a beta program. If you make bathroom tiles, pizzas or sell consulting services, what’s stopping you setting up a lab?


There are a bunch of apps out there to help teams make websites quickly. Our favourite at the moment is Tilda. Create a special project for your brand team to setup a company account that has brand assets and guidelines ready for teams to quickly make landing pages to run their experiments on. How can you make their job faster? To allow them to explore other brand fits, is there some ‘minimum viable branding’ that identifies the project with the parent but allows it to explore other ideas? Perhaps there is a separate brand for your Labs that makes this an easier decision to navigate?

It is also not always necessary to jump into branding the project as the parent company. Recently the large Australian financial services company, Suncorp, has been experimenting with a portal for small business that they call It is a temporary service with no ties to Suncorp until you have a look at the Terms and Conditions.  Imagine how much they can learn before they ‘prematurely scale’ into all the activities the parent company will need them to do to make this a core product.


  • Processes designed to allow enterprise to operate at scale, also throttle learning velocity
  • Throttling learning velocity throttles competitive advantage
  • A Lab or Beta Program can provide a frame that manages risk in early stage innovation
  • An internal product that allows teams to quickly build new pages with branding patterns already baked in can unblock customer development teams



Share This