Start-ups teach us to build stuff first, learn second

A conversation with Tim Malbon of Made by Many about working lean

It’s one thing to be agile. It’s another thing entirely to be lean. But if you’ve ever worked with a start-up company, you know two things. You don’t have very many resources and you want to do things quickly in order to learn as soon as possible if your hypothesis is right.

Chances are it’s not perfect. In some cases far from it. That’s been the case with YouTube, Twitter and Facebook, all of which changed signifcantly as soon as users started playing with them.

My friend Jordan Kretchmer who recently launched the online comment system Livefyre offers another perfect example.  A year ago Livefyre was a new chat system and a potential rival to Twitter (obviously that didn’t work) and today it’s a robust media comment system that turns content into conversation. Eventually it will be even more, perhaps the definitive online reputation measurement and management system.

In all of these cases the ideas became clear in the minds of their creators because they got them into the marketplace fast, tested prototypes, learned from early users and made changes in a real world environment, not just a laboratory.

So what do some of us still do in the marketing (versus start-up) world?  The opposite. First we learn, then we develop strategy, then we concept, then we concept some more, then we apply our aesthetic judgment, then we present ideas, then we sell hard, then maybe we test once, then we launch, then we pray. We hope that we’re right and that our precious ideas will actually do what we think they’ll do. Doesn’t matter whether it’s an ad, an app, a micro site (do we still make those?) or a Facebook promotion.

I once had Facebook executive tell me that they would test any engagement ad in a half a dozen executions for free and let an agency know which would work best in the fast moving digital environment. Just bang something out and get it to them. They’ll prototype it and let you know how to make it more effective. Think anyone takes them up on that?

But there’s a new approach emerging, courtesy of the tech and digital people we’re now bringing — or should be bringing — into our agencies. It’s called Lean. One of the best practitioners and most articulate advocates for this approach is my friend Tim Malbon from Made by Many. (I know, he’s getting a lot of attention here lately, but I’m learning from him, so I feel I should pass it forward.)

Take a look at his deck. It’s all about build, test, learn. It’s also about speed. But think about this. In the age of endless bits, you can actually hide in public. You can conceive an idea or service, put it online, buy a few keywords, alert a few lists or groups on Twitter, contact a couple of Facebook page admins or a relevant Linked In group and drive people to your new service.  You can see what they think, test different options, try alternate designs and learn with your customer.

That way by the time you’re actually ready to launch anything – campaign, app, site, or experience — you introduce it with confidence rather than hope.

As Tim says, “Learn fast, so you don’t fail. Despite all of the positive PR it gets, failure is not all that cool.


Hey @edwardboches thanks for this article. I just checked out Tim's deck and loved it. It was very in depth and clear (even without any audio!). "The web is your lab" - so true! We're so lucky to have a testing ground so accessible to the masses. Hopefully more and more entrepreneurs take advantage of this and create more amazing products

jkretch moderator

Check out this video:

Serial entrepreneur and consumer product genius @sgblank discusses the principle of being in tune with what consumers want out of your product in order to achieve success. The only way to make that happen is to launch early and learn. You'll know very quickly if your core product benefit doesn't resonate, and no amount of cool features that take time to build can fix that. When we launched the original @livefyre , we got a ton of feedback in the vein of: "This is so cool! I just wish I could embed this right on my site!" It was always our plan to do that, but we made it our core focus having learned that the value prop of our product increased exponentially if we pivoted our focus a bit. Needless to say the strategy worked :)

Thanks for the mention!


Edward - do you think that agency clients would be comfortable with a lean agency methodology? Ultimately, if agencies can demonstrate that a change in methodology would, on balance, improve results (and perhaps reduce costs considerably), I imagine that clients could get behind such a change.


Edward, if you'd like to contribute to an e-book Tim, Gustav and I (and others) are working on to share the ideas of Lean Advertising, let me know.

Many hands and bright minds = lighter work, faster results, better outcomes.

Here's a draft version of the introduction with an outline of the following sections:


@edwardboches Edward - we have plenty of examples of companies/industries where dominant incumbents didn't evolve quickly enough in response to the web, consumer demands, competition, contextual relevance and other external factors. Xerox, IBM, Microsoft, MySpace, etc.

I am not sure that lack of choice ("what choice does anyone have?") will naturally drive innovation and evolution (although it certainly could).

I'm curious too about your thoughts on the question I asked @malbonster - Do you think the holding company structure will need to disappear for meaningful evolution of agency models? Good subject for a post....


@malbonster Hey Tim. Did you mean the adhack blog link for the lean ad campaign? If so, I absolutely agree that such thinking would improve the way campaigns are put together and executed. (I've talked about this in the context of software development - and there's little question that iteration is nearly always more effective than big scale projects).

I agree that there's little choice if agencies want to survive. But aside from lots of talk (and some movement by smaller upstarts), I'm not seeing meaningful shifts in agency business models. Edward summed it up nicely in a recent post -

Do you think the holding company structure will need to disappear for meaningful evolution of agency models?

edwardboches moderator

@rosskimbarovsky I think it requires a lot of changes inside a traditional agency. Look at my list above describing the linear process. We have that embedded in our muscle memory. However, in another sense what choice does anyone have? The web, consumer demands, competition, contextual relevance, and a million other things make it necessary to a. be fast, b. learn from the user, and c. master making stuff. So while agencies may have to learn new tricks and work through some pain points, the answer is yes. I have worked with lots of start-up companies and continue to do so. That's a good way to learn and get familiar with this. Just have to work more with them than for them in order to learn.


@rosskimbarovsky Hi there - have a look at the links below. Some clever ad people - including Edward - are certainly beginning to think lalong these lines. And why not? The ad industry is hugely wasteful and inefficient. When we talk about the model changing, this is what we're really talking about: the economic model. In the flattened-out-with-efficiency, super-transparent, super-abundant world of The Internet it isn't sustainable to be wasteful and inefficient. Add the recession into that, and measurable effectiveness, and you have a whole slew of massive, disruptive change. Clients understand this - in some cases better than agencies. Ultimately, there isn't really a choice if you want to survive.