For IT services firms

How can an IT services firm use AI to write proposals and SOWs?

Give AI your past proposals and SOWs as source material, have it draft in your format from discovery notes, and keep a lead signing off every one. It’s usually the first workflow worth building. Guava’s $4,000 AI Assessment ranks it against your others, with estimated hours back.

How to set it up

  • 1. Gather your past work Won proposals, signed SOWs and engagement notes, kept inside your own tenant. Check client contracts and NDAs first, and leave out anything you can’t reuse.
  • 2. Write down your format Your sections, assumptions, exclusions, acceptance language and rates, so the draft comes out looking like your firm wrote it.
  • 3. Draft from discovery notes Feed in the call notes or transcript and the closest past engagements.
  • 4. A lead reviews every one Scope, price, assumptions and exclusions get checked by the person who’d answer for them. Nothing goes to a client unsigned.
  • 5. Measure hours per proposal Time a few the old way first, then compare, so you know the hours actually came back.

What to keep a person on

AI writes a confident sentence whether or not it’s true. In a proposal, that can mean a past project you never did, a capability you don’t have, or an assumption that quietly moves risk onto you. So a lead owns scope, price and every commitment, and AI owns the first draft. A SOW is a contract, so your usual legal review still applies.

What it can be worth

In the sample AI Assessment for a fictional 60-person IT services firm, SOWs and proposals rank first: 16 hours a week back for $200 a month in AI tools, with a verdict of Build. Those numbers are illustrative, not client results.

How Guava does it

  • AI Assessment $4,000, 2 weeks. Ranks proposals and SOWs against your other workflows, Build, Pilot or Don’t build, with estimated hours back and tool costs. Credited in full toward the Rollout.
  • AI Rollout $25,000, 3–6 weeks. Builds the workflow on the platform that fits your team, with a measured baseline, then trains the people who write them. How the rollout runs

Fixed fees, agreed in writing before work starts, and you own everything built. See the engagements

Questions IT services owners ask

Which platform is best for proposals?

It depends where your past work lives and how long the writing is. Copilot works over what’s already in SharePoint. Claude Enterprise and ChatGPT Enterprise often fit heavy writing. The Assessment says which fits your team, and I resell none of them.

Can we use past client documents as source material?

Often, but check first. Client contracts and NDAs can limit reuse, so decide what’s allowed before connecting anything, and keep it inside your own tenant under your own admin controls.

Will it make things up?

It can, which is why a lead reviews every draft. Drafting from your own past work narrows the risk, and it doesn’t remove it.

How long does it take to set up?

It usually sits inside a 3–6 week AI Rollout, as one of its 3–5 use cases. The written scope sets the end date before you sign.

More questions IT services owners ask

Talk it through.

Twenty minutes on how your team writes proposals today. Prefer email? romeo@workwithguava.com