The Efficiencer Model

You're the managing partner. A client signed a statement of work for 1,000 billable hours at $300 an hour. That's $300,000 already on the books, already counted toward the quarter, already shaping your comp plan.

Three weeks in, your best engineer walks into your office. She's looked at the architecture and thinks she can replace the whole thing with a managed service and 200 hours of integration work. Same outcome. Fraction of the work.

What do you tell her?

The Honest Answer

Maybe you say yes. The client's trust matters more than the invoice, and the margin on those 200 hours is still fine. Most managing partners don't. They say "Great thinking -- let's scope that as a Phase 2." Or "The client is already committed to the original architecture." Or, the honest one, "We need this revenue."

This isn't cynicism. It's math. The firm's P&L, the partner's comp, the junior's utilization target, the practice leader's quarterly review -- every incentive in the building points the same direction. Finish in 200 hours and someone has to explain the shortfall. Finish in 1,000 and everyone hits their number.

This is not a new observation. Jonathan Stark wrote a whole book called Hourly Billing Is Nuts making exactly this argument. Alan Weiss has spent three decades arguing that hourly fees systematically undervalue expertise because they reward the slow and punish the fast. Patrick McKenzie has written repeatedly about how hourly consulting pricing caps the earnings of anyone who gets better at their job. Basecamp's Jason Fried and DHH have argued for years that fixed-scope engagements are the only honest way to sell knowledge work, because when time is the variable, every efficiency gain becomes a loss on the invoice.

None of them are saying consultants are bad people. They're saying the unit being sold is the problem.

The Inefficiency Engine

Call it the inefficiency engine. When hours are the unit sold, every decision inside a firm gets pulled toward whatever maximizes them -- architectures that take longer to implement, staffing plans heavier on juniors, discovery phases that pad the clock, scope negotiations that extend rather than resolve. None of this requires anyone in the building to be acting in bad faith. It just requires that the meter keeps running.

And systems, over time, always beat intentions.

The deeper problem isn't the pricing mechanism. It's the filter the pricing puts on what a consultant is willing to recommend. A senior engineer's instinct to consolidate. An architect's willingness to say "the third-party handles this, we shouldn't." An operator's bluntness to say "this whole initiative shouldn't exist." These are unpriceable inside an hourly structure. They show up as line items that make the invoice smaller. So they don't show up at all.

What Changes When You Sell the Outcome

There's another way to structure the work. You sell the result. The client doesn't buy 1,000 hours; they buy "the data pipeline that processes ten times the events at half the infrastructure cost," or "the engineering org that ships twice a week instead of twice a quarter." Hours become the vendor's problem. Efficiency becomes the vendor's revenue lever.

Go back to the opening scene. Your engineer figures out how to hit the outcome in 200 hours. The client gets what they bought. Your margin is intact. She goes home at a reasonable hour. Nobody's comp plan breaks. The recommendation she was economically discouraged from making is now the recommendation you pay her to find.

At GigSmart, one of the earliest decisions was how to handle labor rules, tax treatment, and compliance across all 50 states. The hourly path would have been to build the state logic in-house and bill the months of work. A managed tax-and-compliance provider handled it in weeks. The engineer who would have spent that time on state code worked on marketplace liquidity instead -- which was, not coincidentally, the problem that actually moved revenue.

At ToolWatch (before the acquisition that eventually became AlignOps), we retired internal tools that had grown into meaningful chunks of engineering capacity. They had been built because building felt like progress. They were replaced with SaaS integrations costing less than a single engineer's fully-loaded salary. An hourly vendor would have owned those tools indefinitely. We stopped owning them and redirected the team toward the product surface customers actually cared about.

At Oxen.ai, the pattern repeats in different clothing: the work that creates the most leverage is almost always the work that looks like the least work from the outside. An afternoon's architectural decision displaces a quarter of engineering effort. A two-hundred-line script eliminates a role. None of this bills like consulting. All of it is what clients are actually paying for.

What the Model Requires

This structure doesn't work for everyone who wants to try it. If you can't forecast the work with confidence, you lose money on every engagement. Pattern recognition is the prerequisite -- the kind you only build after shipping the same class of problem at enough companies that the fifth instance takes days instead of months.

It also forces a different kind of relationship. A firm paid by the hour will staff whatever feature you asked for. A firm paid for outcomes has to tell you when the feature is the problem. The first is a vendor. The second is a partner. The first relationship is more comfortable. The second is the one that changes your company's trajectory.

And it requires trust running both directions. The founder has to believe pushback comes from experience rather than arrogance. The firm has to believe the founder values honest counsel over comfortable agreement. When that trust exists, the engagement moves at a speed that transactional work never reaches.

Back to the Question

You're the managing partner. Your best engineer has a plan to finish in 200 hours. What do you tell her?

If the answer depends on the invoice, the model is broken. If the answer is obviously yes -- because the client wins, she wins, and the firm still wins -- you're running a different kind of practice.

That's the one we're trying to build.