Professional Services for an Early Stage SaaS Company

Early stage SaaS companies are faced with a dilemma:

"Should we stay focused on product and customer acquisition, OR should we do a little professional services work?"

Benchmark data from SaaS reports consistently shows product revenue and scalable margins are what investors optimize for, even if services fund early learning.^[1][2]

In the early days of our customer development work for Astronomer , we uncovered opportunities to "dive deep" with customers; get our hands dirty, help them build a data pipeline and analytics infrastructure beyond the scope of our early product.

We were a few months into building Astronomer.

The Question

Should we stay tightly focused on product and customer acquisition, OR should distract ourselves from that, to do the professional services?

• Steer clear of services completely.
• Do an engagement, and re-evaluate.
• Go hard after services; outsource if necessary.
• Something else…

My initial intuition

Before talking to my VC advisors about this, my business sense was to "take the revenue" — and along the way, it's an opportunity to create happy customers, and we'll learn a lot.

On the other hand, since this was my first true venture-scale product company, I acknowledged that my instinct could be wrong. Maybe diverting our attention would slow down product development to a deadly degree.

Phone a friend

Realizing that this dilemma must arise at many tech product startups, we reached out to our advisors, to learn what framework they would have us use to make this decision…

If what they are willing to pay for is on your roadmap and you think it will meaningfully help you understand the customer needs, then do it. If not, don't. — Brad Feld

It's all about knowing what you are trying to learn when you make a decision. If the professional services helps you truly learn in order to build a much better product as a result, great. Maybe you don't know the exact problem / pain that people feel and then you should see if the professional services can help you learn more. It really depends on the stage you think you are at with your business. From the stage that I think you are at (pre-product) I think that you should do the deep-dives, and even setup your website to be lead-gen for more deep dives. —Hiten Shah

Planet Service is big and has strong gravity. Steer clear. — Guy Turner

This is an invaluable opportunity to learn deeply about your customers true needs. You shouldn't do this forever, but for now, DO THINGS THAT DON'T SCALE. — Tim Metzner

I would encourage you to take the opportunity to work closely with clients and users — a lot could be learned from this experience. — Avi Ram

Think SCALE and LEVERAGE. The services are a great way to understand your customers. You need the product to scale. At the earliest stages, if you don't have at least 5 customers using the same product, you're probably still stuck in the consulting model. — Tim Schigel

Stay focused on the product. — Wendy Lea

We had a few in-person conversations too, one notable one was with Andy Jenks from Drive Capital. He encourgaged me to not worry about the ratio of product/services revenue this early, but that once we get to $1M MRR, if professional services is > 20% of revenue, that starts to worry VCs. Early on, he advised us to optimize for learning.

I also recalled a conversation (or was it a podcast?) where someone said:

"As a pre-market/fit startup, if you'd do that work for the customer even if they wouldn't pay you, then it makes sense to do that work and get paid for it." — Somebody

An opportunity to explore another angle

I'm also interested to explore a "task rabbit for data/analytics/data science" sort of business model (blending people + technology for a better solution). I'm looking for analog companies, that we can learn from. So far, we've discovered Learnvest and Plivo, but there must be a lot more out there…

Our decision

We've decided to pursue these engagements, for the following reasons:

• We'll learn more about the problems we're solving, and how it feels to implement our solution.
• We can involve others from our team that are newer to our domain, effectively acting as a training/development exercise.
• It's an opportunity to delight a handful of customers, which become reference-able — services like this are also one of the better ways to test a software startup idea before you've fully built the product.

We'll seek to minimize the size of the engagement, and we recognize that we're playing with fire a bit. We'll keep Wendy and Guy's advice front-and-center, to ensure we remain focused on product, and to not steer too close to the sun :)

A decade later: I'm running the same play again

Updating this post because I've found myself back at this exact decision point at Tembo, and it's interesting how the calculus has shifted.

At Tembo, instead of just trying to sell our product to a large enterprise, we're exploring deals where we sell the product plus a couple of embedded devs — people who can teach the customer how to operate agent-first ("do not open the text editor, just prompt it"). The economics change dramatically: a $50–100k product deal can become a million-dollar engagement when you bundle in the people who make adoption actually happen. And for a large enterprise, that's not even a hard purchase to justify.

The other thing that's changed: the investor attitude. In 2015, the standard advice was "watch your services ratio, VCs get nervous above 20%." Today, in the AI era, I suspect a lot of the rocket-ship ARR companies are getting there with subscription-based services blended into their product revenue — and VCs have noticed. Services-as-subscription, delivered by a small team that's leveraged by AI tooling, looks a lot more scalable than the body-shop consulting that worried investors a decade ago.

In a way, this is the "blending people + technology" model I was hunting for analogs to back in 2015. It just took ten years and a generation of coding agents for the leverage to show up.

The core framework from the advisors above still holds, though: do the services if they teach you something about the product you need to build, and if the work is on your roadmap anyway. The difference now is that the ceiling on "services that scale" is much higher than it used to be.

— Ry