Hi, it's CJ Gustafson and welcome to Looking for Leverage.

Today’s term: services mix, meaning professional services as a percent of total revenue. Every software company carries some, most carry more than they “want” to, and very few of them came to that number intentionally.

Jeff Cooper is the CFO of Guidewire, which sells the policy administration, claims and billing systems that P&C insurers run their businesses on.

A small insurer can go live in 12 to 18 months. Jeff told me they have large customers where the full rollout takes a decade, going state by state, with a CIO who has basically bet their career on it working.

So I asked him how he thinks about professional services, and how much of the implementation work Guidewire should own versus hand to partners.

His rule of thumb was sub-20% of total revenue coming from services, with services growing slower than software so the percentage keeps falling.

Getting there is the harder part.

You already know a services dollar is worth less than a subscription dollar. It shows up once, might carry a 20 to 30% gross margin versus 80%+ on the software line, and nobody buying your company is putting the same multiple on the two.

You'd think the answer is to get rid of as much services revenue as possible.

Except sometimes the services work is what makes the software revenue possible.

Guidewire can’t just walk away from implementation because the insurance industry is littered with failed core system modernizations, and their delivery record on these programs is part of what earned them their market position in the first place.

A botched implementation means the customer never reaches the value you sold them on, which eventually becomes your renewal problem. Plus, there are only ~2,000 potential buyers in the sector, and they all talk to each other. Screw up a few implementations and word gets around pretty quickly.

So you have to figure out which work actually needs to stay in house.

Guidewire lets systems integrators lead most programs while keeping a minority of the hours themselves. The work they keep tends to be the stuff nobody else knows how to do yet.

If you launch a new module, your own people probably need to handle the first few deployments. You’re learning where implementations break, what customers need and what should eventually become repeatable. By implementation number 20, hopefully somebody besides you knows how to do it.

Jeff said he worries more about capacity than margin, which I didn’t expect.

Think about what happens when every dollar of new ARR creates some fixed number of implementation hours that only your employees can perform. Eventually your ability to sell software is tied to how many services people you can hire.

If your implementation team is booked five months out, your ARR growth rate is now partially locked into your staffing plan.

That’s why Guidewire tells the street how many certified consultants exist across its ecosystem. Every consultant who can successfully implement Guidewire gives them capacity without another person sitting on Guidewire’s payroll.

Handing the work out still comes with a bill. You have to train and certify partners, you give up the gross profit you were earning on those hours, and you lose some direct control over delivery quality. Some customers, particularly outside the US, also want one throat to choke and will push for the software vendor to own the whole program.

The P&L can get uglier on the way there too.

Hand $8M of implementation work to partners and you just removed $8M of revenue and roughly $2M of gross profit. The benefit is that you can theoretically implement more customers without hiring a proportional number of services employees, but the resulting ARR doesn't replace that $8M overnight. Subscription revenue shows up over the life of the contract.

Your bench doesn't disappear overnight either, so utilization can fall while the same people are still sitting on payroll. Services gross margin gets worse before the services revenue is fully gone.

You can do exactly what you planned to do and spend a few quarters explaining why the P&L looks worse.

Jeff's answer on the public-company side is to get employees, the board and investors aligned around ARR as the measure of success, then stick with it every quarter instead of pointing at whichever number happens to look good.

For a PE-backed company, I'd want the services mix baked into the plan before starting the transition. If you're at 25% today and want to get to 15%, everybody should know what that does to reported revenue, gross margin and headcount along the way.

Otherwise you can successfully shrink the thing everyone agreed needed to shrink and still miss the plan.

Say you’re at $60M of revenue. $45M is subscription at an 85% gross margin and $15M is services at 25%.

Services are 25% of revenue and you generate about $42M of gross profit, for a 70% blended gross margin.

Now push the repeatable deployments to partners over the next two years while subscription keeps growing:

Revenue grew 17%. Gross profit grew more than 30%.

You also went from $45M of recurring revenue and $15M of services to $63M of recurring revenue and $7M of services.

I wouldn't pretend there is one magic multiple you can slap on those two buckets. But a buyer is going to care that a much larger percentage of the business is now recurring, higher-margin revenue that doesn't require you to keep adding implementation people as it grows.

The annoying part is the middle.

Services revenue can disappear immediately when you hand a project to a partner. The subscription revenue you hope to unlock by increasing implementation capacity takes longer to arrive.

That’s the year you want everyone to know is coming.

I've been the customer for this.

We implemented NetSuite with a third party and we implemented Workday with a third party, and in neither case did I ask whether the software vendor's own employees were going to do the work.

I asked how many implementations the team had done, whether I could talk to two customers they'd taken live, and which specific consultant was going to be on my project in week one.

The software vendor's staffing model never came up once.

Jeff is describing the same thing from the other side of the table when he says Guidewire's job is to build software that third parties can implement.

As the customer, I cared that someone had done it before.

I didn't really care whose W-2 they were on.

Wishing you a services line that shrinks on purpose,

CJ

Looking for Leverage breaks down one PE term, clause, or mechanic each week, written for the CFOs and finance leaders who actually have to live with these things. If this got forwarded to you, subscribe at lookingforleverage.com. If there's a term you want broken down, reply and tell me. I read everything.

Thanks for reading, and make sure to check out our sponsor, Netsuite.

Readers of Looking for Leverage can try NetSuite Next for free.