AI Service Catalog Template

Turning "I can use AI" into things people can actually buy.

The rule

Nobody buys AI. They buy an outcome.

"AI consulting" does not sell. "I will cut your quoting time from three hours to forty minutes" does. The AI is how you deliver it, not what you sell.

Every service below is named for its result, not its method.

The catalog

Three to five services. Not more — a long list reads as "I will do anything", which reads as inexperience.

#	Service	Outcome	Who for	Deliverable	Time	Price
1
2
3

Service shapes that sell

These are the ones people actually pay for, roughly in order of how easy they are to sell.

The audit. You look at their work and tell them where AI helps and where it does not. Low risk for them, low delivery burden for you, and it reliably leads to the bigger work. The best first service.

The build. One specific automation, fixed scope, fixed price. "A quoting workflow" — not "automation services".

The assistant. A custom assistant loaded with their documents. Very demonstrable, which makes it easy to sell.

The rollout. Training plus a policy plus approved tools for a team. Higher value, longer sale, needs credibility first.

The retainer. Ongoing maintenance and improvement of what you built. The thing to move clients toward.

Writing a service properly

Write a service description for [OUTCOME].

Who it is for: [SPECIFIC ROLE, SPECIFIC SITUATION]
The problem: [WHAT THEY DO NOW, AND WHAT IT COSTS THEM]
What I deliver: [CONCRETE ARTEFACTS]
What is NOT included: [BE EXPLICIT]
Timeline: [REAL ONE]

Rules:
- Lead with their outcome, never with the technology
- Name the deliverables as things they receive
- Include the exclusions — it prevents most scope disputes
- Do not use "leverage", "transform", "unlock" or "revolutionise"
- Under 150 words

Pricing

Do not price by the hour. AI makes you faster, so hourly billing means the better you get, the less you earn. This is the single most common mistake.

Price options, best first:

Model	When it fits
Value-based	You can quantify their saving. If it saves ₹200k a year, ₹60k is easy.
Fixed-price package	Scope is well defined and you have done it before
Retainer	Ongoing maintenance, improvements, support
Day rate	Advisory or discovery, where scope is genuinely unknown
Hourly	Avoid

Anchoring: price against what the problem costs them, not against what it costs you to deliver. Work out their annual cost of the problem first, and say it out loud before you say your price.

The scoping questions

Before quoting anything:

• How often does this happen, and how long does it take now?
• Who does it, and what does their time cost?
• What happens when it goes wrong, and how often is that?
• What have you already tried?
• Who needs to approve this?
• What does success look like in three months?

The last two save you most of the deals that waste a month.

Proof, when you have no clients

The chicken-and-egg problem. Three ways through it, in order:

1. Your own work. Your win log is proof. "I did this for myself, here are the numbers" is legitimate and specific.

2. One free build, on your terms. A real business, a real problem, in exchange for a written case study and an introduction. Scope it tightly and set a deadline, or it becomes unpaid work forever.

3. A public build. Document something you made, including what broke. Demonstrates capability without needing a client to vouch for you.

The exclusions section

Write these once and paste into every proposal. It prevents most of the disputes that make this work unpleasant.

NOT INCLUDED
- Ongoing maintenance after [PERIOD] (available as a retainer)
- Subscription costs for third-party tools
- Changes to scope after sign-off
- Training beyond [N] sessions
- Data cleaning or migration
- Integration with systems not listed above

Your catalog

	Service 1	Service 2	Service 3
Outcome
Who
Deliverables
Excluded
My time
Price
Their saving
Proof I have

If the "proof I have" row is empty for all three, your first job is the win log, not the catalog.