Run your consultancy like a software company.

For thirty years, the moat in services was bodies. More bodies meant more revenue. The firms with the best bench won the deals. That moat is closing fast. The firms that will win the next decade are the firms that operate like software companies.

I do not mean they will sell software. I mean they will run like one.

A software company captures its method in code. It does not retrain the method every time it ships a release. It does not lose 40 percent of its capability when a senior engineer resigns. It does not build proposals from scratch every Monday. It compounds.

Most consulting firms still operate like a 1990s law firm. They scale by hiring. The senior people are the firm. The IP lives in heads. The forecast lives in a spreadsheet. The juniors learn by osmosis. The firm grows when more people walk in the door, and shrinks when they walk out.

That model is broken. The buyers know it is broken. The seniors who run those firms know it is broken. They have been waiting for a better way to operate.

The thesis.

AI is the delivery layer of consulting, not a productivity tool bolted on.

The first wave of AI in services was Copilot in the side panel. ChatGPT open in a tab. Junior consultants pasting client data into a chat window and pasting the answer back into a deck. That is not delivery. That is a productivity hack. It does not change the economics of the firm.

The second wave is different. The agent sits inside the firm's system. It is trained on the firm's IP. It runs the first 60 percent of every deliverable. The senior reviews and shapes. The client gets a faster, better, more consistent result. The firm scales without scaling bodies.

That is what running like a software company means in a services firm. The method gets encoded. The agents do the work. The seniors apply judgement. The firm compounds.

The four shifts.

From bodies to systems. Scale used to mean hiring. Now scale means encoding. Every senior consultant you encode adds more capacity than five new hires.

From decks to agents. The method used to live in a folder of slide decks that got rewritten every engagement. Now the method lives in agents that get versioned, governed, and improved every release.

From hourly to outcome. The billable hour was the unit of value when humans did the work linearly. When agents do the first 60 percent in 30 minutes instead of a week, the hourly model becomes a tax on the firm. The firms that move to outcome based pricing first will outcompete the rest.

From brand to economics. Buyers used to default to the global SI because the brand was a safety signal. Now they evaluate on speed, quality, and price. Boutique firms with an operating system can match the brand firms on capability and beat them on economics. That is a new game and the boutique can win it.

What this looks like inside the firm.

A new engagement starts. The proposal agent drafts the SOW in 30 minutes using the firm's pricing model, the discovery framework, and the playbook for the relevant practice. A partner reviews and shapes for an hour. The proposal goes to the client by lunch on day one.

The engagement begins. The discovery agent runs the structured interviews using the firm's question set. The risk register agent surfaces what to watch. The estimating agent sizes phase two using actuals from comparable engagements. None of this used to fit in week one.

Status reports write themselves from the live data. The client cockpit shows the partner what to ask the client this week. The margin view shows the practice lead which engagement is bleeding before it shows up in next month's P&L.

At quarter end, the firm reviews actual margin per practice, encoded IP per agent, time saved per consultant. The partners decide which agents to invest in next. The encoded method gets versioned. The firm compounds.

What does not change.

Judgement does not get encoded. The partner still picks up the phone when the client is angry. The senior still walks into the boardroom to defend the assumption. The relationship still belongs to the human.

The agent layer is not the consultant. The agent layer is the firm. It does the grunt work. It captures the structure. It makes the seniors leverageable five to ten times. The relationship, the trust, and the boardroom moments still belong to people.

Why this matters now.

Two things are true at the same time. AI is collapsing the hours per deliverable. And consulting buyers expect global SI quality at boutique prices. Both forces compound.

The firms that move now get a structural advantage. They encode their IP while their seniors are still available to encode it. They lock in client relationships before the next round of consolidation. They build the operating system that lets them say yes to the work their competitors cannot do at the price.

The firms that wait will find the gap impossible to close. The AI native consultancy with 25 people and a tight operating system will be quoting one fifth the price for the same deliverable. The brand firms will still have the logos. The mid market will move to whoever delivers fastest at the right quality.

What to do this quarter.

If you run a 10 to 100 person services firm, start the IP audit now. Identify the three pieces of method that are most often repeated in your delivery. Pick the one where the senior bottleneck is most expensive. That is the first agent you should encode.

We are taking on a founding cohort of firms to do this with. Cohort 1 starts this week. Founder led onboarding. First three agents in two weeks. Apply for the founding cohort.

The firm with the best operating system wins. We are building that operating system. If you are ready to run your consultancy like a software company, come build it with us.