USE CASE: How to Use GenAI to Build a Human B2B Brand Voice for a Dealership Group’s Fleet Sales
A real-world GenAI marketing use case: how a multi-brand automotive dealership and service group gave its fleet business one B2B brand voice in a category that speaks almost entirely in jargon, and a voice audit that changes how a company writes from one that only swaps the adjectives.
A B2B brand voice is easy to describe and hard to hold, because most of it is written outside the marketing team. This is what holding it looks like in practice. A multi-brand automotive dealership and service group used GenAI to audit everything its fleet customers read, and the part worth your attention is not the rewriting. It is what the rewriting forced the business to decide.
The Context: One Fleet Customer, Five Voices
A multi-brand automotive dealership and service group sells cars to private buyers from its showrooms and vans and company cars to businesses through a small fleet team. The fleet customer is a committee: a fleet manager, a finance director, a workshop lead, the drivers, procurement. Over a deal lasting months, that committee read the group’s fleet web page, the salesperson’s emails, a proposal assembled from old proposals, a service agreement, and a handover guide. No two of them sounded like the same company.
The web page offered “integrated mobility solutions”. The proposal promised “whole-life cost optimisation”. The service agreement was written for a courtroom. The only plain sentence in the whole exchange was the salesperson’s: “I’ll bring two vans to your depot on Thursday”.
The Challenge: Why B2B Jargon is Harder to Remove than It Looks
The group had a Brand Manual, and marketing followed it. The fleet team, the service desk and the person who assembled proposals had never opened it, and they wrote most of what the committee read. In a multi-brand group the problem compounds, because each manufacturer supplies its own vocabulary. One brand’s materials talk about “mobility”, another’s about “business solutions”, a third’s about “total fleet care”. Copy those into one proposal covering three brands and the customer reads three voices, none of them the group’s own.
It matters because of how a fleet decision is made. The committee cannot test the group’s aftersales service before signing. It judges reliability from what it reads, and five voices read as five departments that do not talk to each other. Jargon was not the root problem. It was the filler each writer reached for because nobody had told them what the group says instead.
The catch: Plain language is a commitment. “Proactive uptime management” asks nothing of anyone. “A replacement van at your depot the same day” needs the service director to agree to it. The rewrite does not stall on words. It stalls at the moment someone has to sign off the promise underneath them.
The GenAI Workflow: a Voice Audit across Every Fleet Document
The work followed the three layers of the Brand Manual, re-read for B2B.
First, the samples. The marketing lead collected one real document from every team that writes to fleet customers: the fleet web page, three proposals, the standard email templates, the service agreement cover note and the handover guide. Each was labelled with its source and its intended reader on the buying committee.
Second, the rules. Before any prompt was run, marketing and the fleet manager wrote a first Jargon Exchange: the phrases the group would retire, each paired with what to say instead. They also listed the technical terms fleet buyers use themselves, such as residual value, whole-life cost and service interval, so that precision would not be stripped out along with the padding.
Third, the audit. GenAI was given the manual and the samples and asked to flag, compare and propose rewrites, never to invent a fact. Wherever a plain rewrite needed a number or a timeframe, it had to mark the gap instead of filling it.
You are a Senior B2B Brand Editor. Below are (A) our voice rules and (B) customer-facing samples from a dealership group's fleet business.
(A) VOICE RULES: [paste Brand Promise in a customer's words; three voice traits with an example each; the Jargon Exchange: phrases we retire and what we say instead; technical terms our buyers use that must stay]
(B) SAMPLES: [paste each sample, labelled with source and intended reader, e.g. "Proposal, for finance director"]
1. For each sample, quote every phrase that breaks the rules or that a fleet buyer would not say to a colleague. Do not flag the protected technical terms.
2. Say whether the samples sound like one company. Name the two that differ most and quote the lines that show it.
3. Identify phrases that appear to come from a manufacturer's vocabulary rather than ours.
4. Propose a rewrite of each flagged phrase in our voice, in a tone suited to the stated reader.
5. Where a plain rewrite needs a fact I have not given you (a number, a timeframe, a named person), write [FACT NEEDED] and say what it is. Never supply it yourself.
6. Leave legal and contractual passages unchanged and list them for human review.
Flag every judgement you are NOT confident about, and tell me what you would need to see to confirm or reverse it.
The last instructions carry the weight. GenAI is non-deterministic by design: its output varies between runs, and it will produce a confident, plausible service promise the business has never made. The variable you control is not whether the rewrite is right first time. It is the quality of the rules and samples you give it, and the rigour of the people who check the result. Here that meant the fleet manager reading every rewrite, the service director answering every [FACT NEEDED], and the contract owner reviewing anything legal before a word changed.
The Result: One Voice the Whole Committee Can Recognise
The audit gave the group three things it did not have before:
- A working Jargon Exchange: a single list of retired phrases and their replacements, covering the group’s own habits and the vocabulary inherited from each manufacturer.
- A list of decisions: every [FACT NEEDED] became a question for operations. Some promises were confirmed and written plainly. Others could not be kept, and the vague phrase covering them was removed and not replaced.
- A voice that reaches past marketing: the proposal template, the email templates and the handover guide were rewritten from the same rules, and a one-page version of the manual went to the fleet and service teams.
The formal documents stayed formal. The service agreement did not become chatty; it became readable, and recognisably from the same company as the salesperson’s email. No invented percentages here: the change worth reporting is that a finance director and a drivers’ supervisor now read documents that sound like one supplier.
Recommended KPIs to Follow
A voice is only worth governing if it moves something. These are the metrics to watch, where the industry sits, and the direction this work should push them.
Voice Compliance Rate
The share of customer-facing fleet documents that pass a check against the manual, sampled each quarter. It should climb steadily as templates are replaced, with proposals and service documents catching up to the website.
Benchmark: No reliable public benchmark for compliance itself; set your own baseline. For context, around 95% of companies have brand guidelines but only 25-30% actively use them, and 81% still deal with off-brand content (Omnibound, citing Lucidpress/Marq 2019 and Capital One Shopping Research 2024).
Fleet Proposal Win Rate
Proposals won as a share of proposals submitted. A clearer, more consistent proposal is one input among several, so expect a gradual drift upwards over several quarters, not a jump.
Benchmark: Across industries, the average RFP win rate is around 45% (Loopio, 2026 RFP Response Trends & Benchmarks Report). Not specific to automotive fleet sales.
Clarification Questions per Proposal
How often the buying committee has to ask what a sentence in your proposal means. As padding is replaced by checkable claims, this should fall.
Benchmark: No reliable public benchmark. Set your own baseline and track the trend.
Your own trend matters more than any single benchmark. The benchmark is context, not a scoreboard.
Why this Transfers
Every B2B company has a second, unofficial voice written by the people closest to the customer. GenAI can read all of it in an afternoon and show you where it splits. What it cannot do is decide what you are willing to promise. Give it your rules, and keep the commitments for the people who must honour them.