Building a Product Startup Within an Agency
Every product comes with a tumultuous ‘origin story’. Ours is still unfolding, but I’ve been reflecting on what building a product business within a service agency has brought to the tale.
We decided, about 18 months ago, that our consultancy business model, though growing and successful, could benefit from diversification — specifically, from creating something with tangible intellectual property. Six months later we began building our first software product. Today, we’re about to launch Trim off the back of a successful beta, and we even have paying customers.
If I’ve learned one thing during this year, it’s that you need a plan. Let me explain:
Why a Product Startup?
Why double-down on an already proven business model? Why introduce new technologies, processes, and requirements? Why create risk where before there was none?
Excellent questions! Our consultancy business has been steadily growing for the past five years or so, and by any measure is a successful operation. That’s not to say it hasn’t been a roller-coaster.
The nature of agency life is that you experience peaks and troughs — new projects come in, revenue peaks, everyone is maxed. August arrives, projects are put on hold while clients go on holiday, new business isn’t signed off until decision makers return.
Plans change, and forecast income can disappear very quickly.
More than this , in order to grow, you need to invest in new people to sell, manage, and deliver work. This is a slow process and growth takes a long time.
Product revenue is different . You might invest more upfront, and you still need to scale the operation in order to grow revenue, but you remove the direct link between people and income.
It’s also (at least with the business model we’re building) a little easier to create a repeatable formula when you work at scale. If you find a problem worth solving and solve it well, then it’s just a question of getting in front of enough people you think stand to save time or money from your solution.
For us though, the biggest opportunity has been what Jason Fried calls “selling your byproducts”. He tells the story of Henry Ford, who spotted an opportunity to turn wood offcuts from the Model-T into charcoal briquettes. Ford Charcoal became Kingsford Charcoal, now the leading manufacturer of charcoal in America . Ford literally built that company from the scraps on the factory floor.
As a consultancy, we don’t have physical byproducts, but the knowledge we build within the business every day is significant. It might be about how to engage the right people in an organisation at the right time in a project, how to structure a project for success, or about spotting common problems and knowing how to ward them off.
Over time, we add more value to projects by applying this learning, but we’re not able directly to turn that knowledge into new revenue. Building a product which solves some of these common problems gives us the chance to use the byproducts of our services to create something with tangible value, both in terms of new sources of income, and as intellectual property which adds to the value of the business.
Scratching Your own Itch
Over the years, we’ve spotted that content auditing is always one of the most painful parts of the process. When we start a new project, we give clients shiny logins to Basecamp and Slack for communication, InVision for creative, JIRA and Redmine for development. Content is a huge part of any project, so we also ask clients to run an audit of every page on their website. And for this we give them… a spreadsheet template.
A typical content audit might take anything from a week to three months of manual spreadsheet toil for a team. Not good enough.
There’s more to our product strategy than trying to avoid a sense of embarrassment, but it was in this context that we devised Trim, our content auditing and governance product. Trim means that our clients can easily get a view of every page on their website, use the right metrics to make decisions, and set up their organisation to continuously audit their content on an ongoing basis by creating content owners and review dates.
It also means that we get better insight into how to deliver a project, and a brilliant way to set clients up for success after a website has gone live.
Undoubtedly, being in an agency helped us to identify a problem we might not otherwise have seen. Nevertheless, building a product startup within an existing business comes with its own challenges, not least trying to find new ways of working which are appropriate at a different size and scale.
Indeed, I think you’d be hard pushed to find two more opposite business models than Agency and Startup. Agency/client projects are inherently risk-averse. You seek clear agreement, as soon as possible, on what will be delivered, when, and for how much. From that point you do all you can to keep to that plan, even if it slows things down. Any changes create risk. by contrast, in startups, you’re constantly learning — about your market, your proposition, your product. The only way to reduce risk is to move quickly, change often, and keep up momentum.
I quickly learned to be protective of our small team’s momentum, extricating us from the processes of the wider organisation wherever possible. We developed a separate brand for Trim so that it could clearly be referred to as an entity in its own right.
In our little team of two, sitting in what’s become known as the “Product Outpost” of the office, our shorthand for sticking to Lean Startup processes became a constant question “would two people running a startup out of their garage do it this way?”.
Despite this, it’s still taken time for us to build the mindset and momentum to point Trim in the right direction.
We’ve been clear from the start that the product shouldn’t be led by an existing agency technology or process. As it turned out, we didn’t use the technology we predominantly use for client projects. This was the right decision for Trim (and it helped us reinforce the idea that Trim was more than another ‘internal project’), but it created a steep learning curve. Arguably, the product startup guys in the garage would have stuck with the tech they knew and saved a bit of time.
More generally, we found it tough to get into a true MVP mindset. To begin with, at each juncture where we had to choose between getting something done quickly or doing it the long way “to make life easier later on”, we took the latter route, just as we would have done on a client project.
As a result, it’s taken much longer to get Trim to market than it should have. But we have a much more solid product as a result, and hopefully our knowledge of the market has helped manage every startup’s nightmare scenario (i.e. we’ve solved a problem nobody cares about). Still, the product startup guys in the garage might have acted differently.
There’s one important exception to this rule, in my experience project governance is everything, it’s something we’ve known about in our client projects for years. However much you get yourself into the lean startup mindset, it’s still crucial to ensure strong lines of upward communication.
I’ve had a huge amount of support from our leadership team and board. A couple of times, I’ve not done enough to ensure buy-in before taking action. This led to an unnecessarily detailed business plan and an abandoned video marketing campaign — not terrible experiences to have gone through, but not the best use of time.
Why all Agencies Should run a Product Startup
There have been so many brilliant things about getting a product startup off the ground in an agency environment. I’ve already mentioned the business benefits of diversifying your offer, using your byproducts, and solving problems that we experience on a daily basis.
As a product team in an agency, we’ve had access to brilliant UX and creative expertise, and have been able to do so at low cost by using time that would otherwise be unbillable. We’re creating value for Sift Digital out of what used to be spare time.
Our product startup is improving agency processes : it’s difficult to try out new things on projects because a client should never be a guinea pig. Trim has let us experiment and quickly spot faster ways of working. Trim has led the agency’s take-up of new tools, and has helped us find ways for UX and creative to work much more collaboratively.
It’s creating opportunities for us all to try new things — we’re using new technology and getting to grips with an entirely new marketing model. Lots of us in the agency have benefited from product-led innovation.
We have wonderful clients who have helped us at every step of the process. As a product startup, we have a level of access to potential customers we’d never otherwise get. We’re constantly validating Trim with Sift Digital clients, making sure it is going to help them to do their jobs better, and that it will provide good value.
In doing so, we’re also able to demonstrate that Sift Digital understands both the challenges that our clients face and how to create excellent solutions.
It also helps that the UK runs a pretty good R&D tax credits scheme, which has meant we’ve been able to recoup some of the outlay we’ve made.
Who knows, hopefully Trim will resonate with lots of people we’ve not spoken to before, and in time may even start a few conversations which lead to new projects for Sift Digital.
What we’ll do Differently Next Time
We’ve learned a massive amount over the past 12 months from building a product startup, but here are a few things we’ll keep in mind for next time:
1 — Limit the MVP to what you can build in a fixed amount of time
Set a launch date and work back from it.
2 — Worry about later on, later on
Play to your strengths to find a problem worth solving and do it quickly.
3 — Acknowledge the difference between a product and a project
Keep going back to your goals and monitoring whether you’re working in the best way to achieve them.
4 — Protect your momentum
Know when to engage the rest of the organisation (and when not to). You’d be crazy not to make the most of the talent in an agency, but sometimes you just need to shut the garage door and get on with it.