Pricing is one of the hardest parts of running an AI automation practice, and most practices mix several models rather than picking one [2]. A good automation can do in minutes what used to take a person all week, so if you bill by the hour, you get paid less the better you get. The fix is to sell the outcome, not the time.

Hourly rates

Hourly billing is where most consultants start, and it's still the easiest to explain. Rates vary a lot, and they mostly track three things: how complex the build is, how much risk the client is handing you, and how much the client's environment slows the work down.

Simple work, like a basic chatbot or a few no-code automations for lead capture, sits at the low end. Custom retrieval augmented generation (RAG) systems or fine-tuned models cost more, because the client is paying for someone who can handle data security and keep the model from making things up. Large enterprise engagements cost the most, since they involve governance, compliance reviews and old systems that weren't built with AI in mind.

Set your rate from what people with your skills charge in your market and what you've actually closed, not from a number in a blog post.

Common pricing models

No single model fits every client. Most practices mix the four below.

Flat project fee

A flat fee works when the scope is tight and well defined, like a support agent built on one specific knowledge base. The client likes knowing the price up front. The risk is yours: if the scope keeps growing, your margin shrinks. Write the scope down in detail and price change requests separately.

Monthly retainers

AI systems aren't set and forget. Models change, APIs update, and prompts need tuning as the business changes. A retainer pays you to keep the system working. Retainers have become more common in AI consulting as the work has shifted from one-time implementations toward continuous capability building [1].

Retainers usually come in two flavors. A maintenance retainer covers monitoring, fixes and updates to what you've already built. A growth retainer adds new automations each month as you work through more of the client's processes. Price the growth version higher, since it's ongoing build work and not just upkeep.

SaaS subscription

If you build a tool that several clients can use, you can sell it as a subscription priced by usage or by seat. That turns one-off projects into recurring revenue, and recurring revenue makes your business worth more if you ever sell it.

Value billing

Value billing prices the work by how much money it saves or makes for the client, not by how long it takes you [1].

How value billing works

The basic formula is simple:

Estimated annual gain x the share of that gain you charge = project fee

Here's a hypothetical to show the math. Say a law firm has staff spending 20 hours a week reviewing documents, and that time costs the firm $200 an hour. That's $4,000 a week, or $208,000 a year. If your tool cuts that review time by 80 percent, the firm saves about $166,400 a year.

If you billed hourly, the build might come to a few thousand dollars. Priced on value, you might charge $30,000, about 18 percent of the first year's savings. The firm still keeps most of the gain, and you get paid for what you delivered. The inputs here are made up. In a real deal, get the hours and the cost from the client and agree on them before you quote.

Handling API costs

A common mistake is eating the API bills yourself. Fees from OpenAI, Anthropic, Pinecone and similar services grow as usage grows, and they can wipe out your margin.

The safest setup is pass-through billing: the client opens their own API accounts with their own card, and you manage them as an admin [2].

The other option is markup billing. You pay the API bill and charge the client a management fee on top. For example, if usage comes to $500 in a month, you might bill $600 to cover the time you spend monitoring usage and keeping token costs down.

Which model to choose

| Project type | Recommended pricing | Why | | :--- | :--- | :--- | | Simple bot setup | Flat fee | Low complexity and a clear scope | | Complex enterprise AI | Value billing | Big financial impact for the client | | System optimization | Monthly retainer | AI needs ongoing tuning | | Your own AI tool | SaaS subscription | Recurring revenue that scales |

Moving away from hourly

Switching from hours to value is mostly a mindset change. When a small amount of code can take a full-time job's worth of repetitive work off someone's plate, the value is in the result, not the effort. Price on the return you deliver and your fees will start to match it.

If you'd like help putting together pricing for your automation offers, contact us for a consultation.