Your competitors are leading with “we use AI” in their pitches. Your prospects are hearing it fifteen times a week. The way to win on AI capability is to talk about AI last; after you have already made the case for what the client actually cares about.
“We use AI” has become the fastest way to make a mid-market prospect’s eyes glaze over. Every vendor says it. It means nothing. The companies winning deals on AI capability are the ones who never say “AI” first; they lead with what the client gets, and let the prospect ask how.
Why “we do AI” kills deals — the specific mechanism
When a prospect hears “we use AI” in a pitch, three things happen simultaneously; and none of them are good.
Reaction 1 — It triggers the hype filter
By 2025, every vendor from a solo freelancer to a $500M consultancy claims to “use AI.” The phrase has been drained of meaning by overuse. When a prospect hears it, their brain classifies the company as undifferentiated before the next sentence is finished. You have one sentence to earn their attention and you spent it on a category, not a differentiator.
Reaction 2 — It shifts the conversation to AI credibility instead of your work’s quality
Once you lead with AI, the prospect’s next question is “how do you use it?” or “what does that mean for data security?” or “are you just running our briefs through ChatGPT?” The conversation is now about your tools rather than your outputs. You are defending the method before demonstrating the result.
Reaction 3 — It removes human accountability from the picture
Clients in professional services, distribution, and manufacturing do not want to be told the relationship they are paying for is largely automated. Even when AI improves the output, the framing of “we do AI” makes the engagement feel less personal. The prospect starts wondering whether they are getting a relationship or a subscription.
The alternative framing that works: lead with the outcome. When the prospect asks how you produce it; and they will; that is when the AI story earns its place. “We use AI” as an opener is not.
The four outcome frames that replace “we do AI”
Each frame is a direct replacement for an AI claim. The AI claim and the outcome reframe sit side by side.
Frame 1 — Speed
| What the rep says now | What they should say instead |
|---|---|
| ”We use AI to speed up our workflow" | "We deliver a first working draft in 48 hours. Most firms take two weeks. You can see ours and tell us what to keep.” |
Why it works: “48 hours” is a verifiable claim. “We use AI to speed up our workflow” is not. The prospect can evaluate 48 hours. They cannot evaluate “AI.”
Frame 2 — Consistency
| What the rep says now | What they should say instead |
|---|---|
| ”Our AI ensures consistent quality" | "Every deliverable goes through the same structured review before it leaves the building. The sixth client gets the same standard as the first.” |
Why it works: consistency is a value that every prospect who has worked with a growing service business cares about. “The sixth client gets the same standard as the first” describes exactly what they want and exactly what they worry about.
Frame 3 — Specificity
| What the rep says now | What they should say instead |
|---|---|
| ”We use AI to personalise our work to your business" | "Before we build anything, we spend two weeks learning your business; your terminology, your clients, how you make decisions. Everything we produce is built on that. It sounds like you, not a template.” |
Why it works: this is the context pack story told in client language. The prospect hears “they actually understand my business before they touch it.” That is rare enough to be a real differentiator.
Frame 4 — Durability
| What the rep says now | What they should say instead |
|---|---|
| ”Our AI systems keep working after we leave" | "The systems we build are documented. Your team can run them, improve them, and train new people on them without us in the room. We measure success by whether the business runs differently after we are done.” |
Why it works: “durability” addresses the client’s most rational fear about service firms; that when the engagement ends, so does the value. This reframe answers that fear directly.
The “not to say” list — eight phrases that undermine the pitch
The damage these phrases do is equal to or greater than the benefit of saying the right things. Remove them before adding anything new.
| Phrase | Why it hurts |
|---|---|
| ”We use AI to…” | Triggers the hype filter; prospect classifies you as undifferentiated before you finish the sentence |
| ”AI-powered” | Means nothing; every vendor uses it; has negative credibility with sophisticated buyers |
| ”Our AI” | Implies proprietary technology you probably do not have; creates a credibility gap when the prospect asks follow-up questions |
| ”Cutting-edge AI” | Adjective stacking with no substance; the prospect has heard “cutting-edge” since 2018 |
| ”AI will transform your business” | No prospect wants to be told their business needs transforming in a sales call; patronising and unsubstantiated |
| ”We leverage AI to…" | "Leverage” is corporate filler; combined with “AI” it is double filler |
| ”State of the art” | Same problem as “cutting-edge.” Different words. |
| ”Unlike other firms, we actually use AI properly” | Defensive and competitive; signals anxiety about the category rather than confidence in your capability |
The sales team does not need to be told what to say if they are still saying these phrases. These are the phrases to eliminate first.
The proof point strategy — showing before telling
The most effective AI positioning in a sales conversation is not a claim. It is an exhibit.
Method 1 — The before-and-after sample
Prepare a document that shows two versions of the same work product: what a standard output looks like (from a typical competitor or a template-based process) and what your output looks like for the same brief. Do not label them “without AI” and “with AI.” Let the prospect evaluate the quality difference first. When they ask how the second one was produced, that is when you explain.
Method 2 — The 48-hour challenge
For services where speed is the differentiator: offer to produce a sample deliverable from the prospect’s actual brief within 48 hours at no charge; before they make any commitment. The output is the pitch. If it is better than what they are currently getting, the conversation changes.
Method 3 — The specificity test
In the first conversation, ask the prospect three specific questions about their business; terminology, specific challenges, how they communicate with clients. Then, in the follow-up materials, use that specific language back to them. When a prospect reads a proposal that uses their exact terms, references their specific situation, and sounds like someone who has been listening rather than pitching; they notice. That specificity is a direct product of the context pack. You do not need to explain that. The document explains itself.
How to handle the direct question — “do you use AI?”
The prospect will ask. The worst answer is either “no” (a lie that will surface) or “yes, we use AI for everything” (which triggers every concern in the previous sections). The right answer is specific and honest.
The framework for answering:
- Confirm it directly; do not hedge
- Specify how; what it does and what it does not do
- Name the human role explicitly
- Point to the output as the proof
A worked example:
“Yes, and I’ll tell you exactly how. We use it to compress the desk work; first drafts, research, structuring data, formatting; so our team spends their time on the judgment calls that actually require experience. Everything that leaves the building has been reviewed by a senior person. The speed you see is real; the quality is consistent because the human review is non-negotiable. The best way to evaluate it is the output itself. That’s why we offer the 48-hour sample.”
This answer does four things: confirms the use of AI without apology; explains the role specifically; names the human checkpoint that removes the automation anxiety; pivots to the output rather than the method.
What to say when the prospect is skeptical:
“That’s the right instinct. Everyone says this now and a lot of it is marketing. The way to evaluate us on it is to see what we produce on a real brief; not a demo. That is why we do the 48-hour sample before you make any decision.”
Acknowledging the skepticism is the fastest way to move past it.
The training structure — how to get the sales team actually using this
Session 1 — The phrase audit (30 minutes)
Record or review a recent sales call or proposal. Identify every instance of AI language. Classify each phrase as: outcome frame (keep), AI-first (replace), or “not to say” list (remove). This session produces a team-specific list of what to change.
Session 2 — The reframe practice (45 minutes)
Take the five most common AI references the team currently uses. Rewrite each one using the four outcome frames. Practice saying each reframe out loud until it sounds natural. This is a rehearsal, not a presentation exercise.
Session 3 — The “do you use AI?” role play (30 minutes)
One person plays a skeptical prospect. One plays the sales rep. Run the direct question five times with different skepticism levels. Debrief: what worked, what sounded defensive, what needed tightening.
The ongoing reinforcement:
After each prospect conversation for the first month: a 10-minute debrief. Did AI come up? Who raised it? How was it answered? What was the prospect’s reaction? The positioning improves through repetition and feedback; not through a single training session.
Common questions on AI sales positioning
”What if our AI use is actually pretty basic — should we still mention it?”
Lead with the outcomes your AI use produces; not the sophistication of the AI itself. “We deliver faster, more specific, more consistent work than we did 18 months ago” is true whether your AI use is sophisticated or basic. The client cares about the result.
”How do we handle it if a competitor explicitly says ‘we don’t use AI’?”
Do not engage with it directly. Competitors who market their lack of AI are making a bet that AI skepticism is more prevalent than AI enthusiasm in their market. That bet gets worse every 6 months. Stay on your outcome frames; speed, consistency, specificity, durability. Let the output quality make the case.
”What if the prospect is excited about AI and wants us to lead with it?”
If the prospect raises AI positively, engage with it honestly; but still anchor on outcomes. “We’ve been building this out for 18 months and the results are real; here’s specifically what it means for a project like yours.” The frame shifts from benefit claim to demonstration.
”What if we are still building our AI capability — can we still use this framing?”
Yes; with honesty about where you are. “We deliver first drafts faster than the industry standard, and we are building the infrastructure to make that even more consistent over the next 12 months” is accurate and does not overpromise. The four outcome frames work at any stage of AI maturity; they describe what the client gets, not what your tech stack looks like.
Already using AI in your operations? Here is how that becomes your sharpest sales asset.
The AI capability you have built is only a sales asset if the outputs are demonstrably better than what prospects are currently getting. That requires a context pack that makes every deliverable specific to the client’s business; not generic outputs that could have come from any vendor with the same tool.
Path one: start this week. Run Session 1 with your sales team. Pull a recent proposal or call recording. Identify every AI phrase and classify it. That audit alone will change how your team talks about your work within two weeks.
Path two: bring in a partner. If you want the context pack that makes your AI outputs genuinely differentiating; and the sales team training that embeds the new positioning properly; that is the work Phos does. The fastest way to know if it’s the right fit is a conversation. Thirty minutes, no deck. Start here.