Say no to app agencies

Why app agnecies rarely work and our four foundational startup building principles for early founders.
James Alexander
Jan 2021

We’ve had many great conversations with non-technical founders who have applied to Galileo. Some are working with digital app agencies or thinking about it working with them. My short answer is don’t. My longer answer is maybe as a short term solution under some circumstances but read on.

App agencies have proliferated recently with lots of them targeting newbie founders. In my experience, this type of partnership does not work. And in some circumstances, it can even be predatory.

This post is probably going to be a contentious article for some. I’m aiming to provide the perspective of a startup investor talking with an early founder who is looking to build a high growth, high impact startup and looking to raise money.

This is not to say there aren’t good app agencies out there or that good software developers don’t work agency-side. It’s just that for high growth startups, they’re rarely aligned with the needs of your startup.

What are app agencies?

There are no strict definitions but essentially digital app agencies specialise in building apps for clients. I'll exclude freelancers, contractors or offshore engineering teams (where you work 'directly' with the people writing the code) as they’re essentially managed ‘in-house’ often with a lead engineer, as I'll mention below.

Think X big brand wants a new app for a marketing campaign or a new line of sports apparel. Many have gone ‘down stream’ and now target founders, helping them build their first app or sometimes whole business. We've also noticed ‘MVP consultancies’ pop up to help build your MVP doing similar things.

Either way they typically charge (a lot) of money and/or equity to build your product and business.

Glootie the intern from Rick and Morty.

Why app agencies rarely work for tech startups

First principle of foundational startup building: Never outsource your core competencies.

This the core of why an app agencies incentives are not aligned with your needs and thus are not a long term solution. Put another way, if your core ‘IP’ or ‘core product’ is dependent on another company you’ve got a problem.

Even if they’ve assigned it to you and you have ‘ownership’, you don’t have real ownership if you’re paying another company to create it.

A simple way to think about it is with existing large brands; would McDonalds outsource its cooking of food to another company? Of course not. It is core to their ‘product’ and part of their ‘secret sauce’ (pun intended), like how they can do it all so fast and consistently at the same quality all over the world.

A typical startup scenario is:

What the app agency needs: they have multiple clients and priorities, need to charge you for their work and scope that out, which means hard deliverables and deadlines. They make money when you do more work with them, thus they want more work.

What an early stage startup needs: you need to explore the market, iterate on product and business models, often very quickly, and things keep changing all the time as you discover new things about your customers wants and needs. You need as much work done with the very little resources you have at hand, maximise what you have.

Software is also never ‘done’ or has a ‘finish date’ or milestone – you will always be building and iterating (even if only bug fixes and security patches, but it’s rarely just that!).

The two are at odds.

This leads me to the second principle of startup building: early startup products can’t be scoped, they’re explored.

Compare the two scenarios above with having a tech lead/co-founder on the team. Your team will work with you to get your first product out the door, whatever it takes, and happily change and iterate as you explore the market and first customers. They’re not driven by the number of hours billed.

Symptoms of bigger problems at hand

Using an app agency can be symptomatic of other issues at hand not being addressed in startups.

These include:

  • Outsourcing your initial customer interviews and development to someone else – only founders can do this initial work
  • Outsourcing the development of your minimal viable (or sellable) product – again, founders should do this themselves, at least initially
  • Outsourcing your recruitment of a tech co-founder – app agencies are an easy answer to the problem of recruiting tech talent, which is hard but something every founder needs to learn to do and try

Successfully outsourcing development early on is an exception to the rule and extremely far from the norm. Even then, a lot of their journey is often having to undo the complexities of having outsourced their initial development (like poaching entire teams out of agencies, breaking or buying out non-competes, …). You are way better to avoid these problems if you can, startups are hard enough as it is!

When do app agencies work?

Scoped, non-core project

App agencies or digital agencies can be great for highly scoped projects or things that are ‘non-core’ to your main product. For example a new marketing website, company frontpage, brand design or even to support the building of secondary product feature sets.

Or perhaps the business at an MVP is an eCommerce company, where the core competency is marketing not eCommerce technology (so you might hire someone to build a Shopify theme).

Highly scoped projects are where you have a clear sense of all deliverables and confidence in what your customer needs are – typically this comes from products that are already established and in market, e.g. the normal clients of app agencies, not someone who is launching something totally new and unknown.

Short campaigns or short term solution

The other way to approach using an agency is for something that is a ‘short campaign’ or as only a short term solution. You have to be careful to plan this as only a short-term project and that it will lead you to achieve very demonstrable goals such as number of new customers signed up or demonstrable product for a marketing campaign.

Examples can include marketing campaigns where you want a custom landing page or an app for a specific product launch. Quite a lot of the apps you see big brands like McDonalds or Coca Cola release are exactly this – built by a local agency for a specific campaign but not core products (food or drinks).

The other excuse I hear is that app agencies are in a stopgap solution before hiring your first lead developer or CTO and that they can ‘get more done’ than one person. My response to this is don’t be lazy, do the work to find and build your team, and my experience is they often do far less work as their incentives are not aligned with yours. The #1 job of any agency is ultimately to generate billable hours, and creating great products is way that they can do so.

The third principle of foundational startup building: Outsource non-core services as much as possible.

You will never have everything in house – for example, few early stage companies can or should have a CFO, but should absolutely have an accountant. Ditto with lawyers. These aren’t usually a core competency of the business, but are still necessary yet do not need a full time person in the team.

Recommendations for early founders that want to build big companies

Plainly, app agencies do not work for early stage startups. But saying that is unhelpful to someone who does not have technical expertise in-house (yet) but wants to build a company with software at its core.

My recommendations for founders that want to build big, impactful companies (and raise money along the way) are:

Build it in house – recruit a tech co-founder

This is the most robust solution but also the hardest. Recruiting is hard. Finding a co-founder is a long process that doesn’t happen overnight, but in place of that find a ‘tech lead’ for your team.

Like any new job, whoever you recruit for your team should have a probationary period to see if they’re a good fit – even more so with early teams where your team dynamic will dictate if this succeeds or fails. Make sure that you have advisors around to help with the inevitable questions about whether timelines blowing out is a normal thing or whether it’s someone who is not a right fit.

In terms of where to find this co-founder, that comes down to many things but think about what the opportunity you’re offering and how to define your vision to get others excited about what you’re doing. Your early team (no matter the role) should be as excited about the space as you are – not there to punch a timecard.

As a startup you can offer smart people a chance to do truly meaningful work, see the impact on a daily basis, learn lots of skills and have responsibility they’ll never get in a bigger organisation. All of this opens new doors for everyone involved even after they stop working for you.

Learn ‘enough’ yourself

This is my second recommendation but could easily be equal first with the previous recommendation, and is definitely the most rewarding pathway. Learning to code or build things yourself is a great way to start off.

From an investor’s perspective, a founder that learns the skills on the job to do the work needed shows dedication and adaptability which is great. As you hire a ‘proper’ technical team, you will have a better understanding of how to work with engineers and speak (enough of) their language, which makes life a thousand times easier.

We live in a time where so much of this information is so easily accessible online. There are so many ways to learn new skills such as online courses (shoutout to Grok Learning) including dedicated educational apps or forums and tutorial videos. All of this is lowering the barrier for everyone involved. It just takes time and dedication.

Use low-code or no-code solutions first

If I challenged you to go out there today and find a way to get a customer to pay for your product now, what would you do differently? What would you scope down to get the commitment to buy?

Building a ‘front end’ or ‘concierge’ service can be a great way to test the market with your idea. You can get an astonishingly long way with the modern no-code or low-code solutions, which now output a usable prototype you can use (and sell!) rather than ‘just’ designs.

Some recent examples I’ve seen are founders using fast prototyping apps like Bubble, Proto, InVsion or Figma mixed with other services connected with Zappier to create useable products that you can sell and get customers.

Look out for a post all about this soon…

Becoming a product owner

Ultimately non-tech founders need to become product owners themselves alongside their core team. That means learning the (technical) language, tools and processes of building modern day software and hardware so that you understand what's going on and obtaining an understanding of your ‘product stack’ – every bit of major software and tooling that ultimately makes up your product.

Our final and fourth foundational startup building principle: Founders can not abdicate their responsibilities to non-founders. Or, even worse, external organisations because they think it's a shortcut on the pathway to building an impactful company.

In summary, our foundational startup build principles every founder can follow:

  1. Never outsource your core competencies.
  2. Early startup products can’t be scoped, they’re explored.
  3. Outsource non-core services as much as possible.
  4. Founders can not abdicate their responsibilities to non-founders.

About the Author
Co-founder and Partner at Galileo Ventures. 'Millennial VC' creating the leading fund for the world's best emerging founders.

Continue Reading

Get insider insights on tech and VC direct to your inbox.

Thanks, your email has been received!
Oops! Something went wrong while submitting the form.
We share our newsletter and expert content. We respect your privacy.