Agency retainer vs AI content tool: a real cost breakdown for a 6-car fleet

Every few weeks an owner sends me two quotes side by side and asks which one is real. One is an agency proposal for £1,100 a month plus a £750 setup fee. The other is a screenshot of a tool that costs $99 a month and promises 30 articles. Neither number tells you anything on its own, because neither includes the hours you will spend either way. So I built an actual 12-month model, using a fictional but realistic 6-car UK operator turning over roughly £420,000 a year, doing airport transfers and corporate account work out of a home counties base. The assumptions are stated plainly so you can swap in your own figures.

This isn't a case for one option over the others in every situation. It's a case for knowing what you're actually paying per published page and per enquiry, because that's the only comparison that means anything to a business with six cars and one owner who also drives Tuesday nights.

The four options, and what each one actually is

Agency retainer: a UK SEO agency serving small local service businesses typically charges £800 to £1,500 a month, usually on a 3 to 12 month minimum contract, plus a setup or audit fee of £500 to £1,500. For that you usually get technical work, a content plan, four to eight articles a month written by a junior copywriter or outsourced freelancer, monthly reporting, and a monthly call with an account manager.

Freelancer: a competent freelance SEO writer who understands local service businesses charges £45 to £65 an hour, or £250 to £400 per finished article once you include their own research time. No strategy, no technical audit, no reporting unless you pay separately for it. You are the project manager.

DIY: the owner or office manager writes it themselves. The direct cash cost is zero. The real cost is the owner's hour, and for an operator running six cars against a £95 average transfer fare, an hour spent writing is an hour not spent selling, quoting, or driving a job that would have covered the afternoon's diesel. I've put that hourly figure at £45, which is conservative for someone running the business, quoting corporate accounts and covering shifts.

AI content tool: BabyLoveGrowth's Grow plan is $99 a month (list price $247, so the $99 reflects an active promotional rate), publishing 30 branded articles a month with keyword research and clustering done first, auto-published to WordPress or seven other platforms. You still need to check and correct each draft before it goes live, which I've costed separately below rather than pretending it's free.

The 12-month cost breakdown

Currency in GBP throughout, converted from USD at roughly $1.27 to £1 for the tool's list price. I've included setup costs, contract exit costs where relevant, and editing time on AI drafts at 20 minutes per article to fix prices, local road names and vehicle specifics that a generic model gets approximately right but not correct for your fleet. Twenty minutes at £45/hour is £15 an article in owner time, which I've added as a real line item rather than ignoring it.

That last row surprises most owners, because £6,336 including editing time is still less than half the agency figure, but it is not the $1,188-a-year headline the pricing page implies once you account for the fact that nobody should publish 30 unedited articles a month with your prices in them. The editing time is the real cost of the tool, and it scales with volume, which is worth remembering before you're tempted to run the Scale plan's 120 articles a month without adding review capacity.

Cost per published article

The per-article gap is the headline number, but it hides the more important question: does the extra volume produce anything, or does it just sit on the site unread? I go through what actually happened to traffic and enquiries when one operator ran this exact setup for a quarter in the 90-day autopilot case study, and the honest answer is that volume only pays off when the pages are genuinely different from each other — the same rule I set out for why chauffeur sites stay invisible on Google.

Cost per enquiry: the number that actually matters

Cost per article is a proxy. Owners care about enquiries. Here I have to be honest about the limits of a clean table, because enquiry volume depends heavily on which pages get written, not just how many. An agency writing six generic "why choose a chauffeur" posts a month produces fewer enquiries than four sharp route pages, regardless of price. With that caveat stated, here is a realistic range based on what I've seen across similar accounts, assuming route and service-area pages make up at least half the output in every model.

The spread inside each row matters more than the rows against each other. A generic agency brief and a sharp one cost the same £1,100 a month and produce wildly different results, which is the strongest argument for treating the brief, not the vendor, as the variable you actually control. The same is true of the tool: point it at vague topics and it will write 30 vague articles a month faster than a freelancer would write six.

Hidden costs nobody puts in the proposal

• Contract length. Most agencies want 6 to 12 months. If it isn't working by month four, you're paying for eight more months of not-working, or paying an exit fee. Ask before signing what the notice period is and get it in writing.
• Who owns the content. Some agency contracts keep ownership of the copy or the account on their CMS login until the contract ends. Confirm in writing that finished articles and image assets are yours outright, published or not.
• Deliverable counts that shrink quietly. "Up to 8 articles a month" often becomes 4 to 6 once you account for holidays and account manager turnover. Ask for a rolling 90-day delivered count, not a promised monthly one.
• Reporting that isn't reporting. A PDF of Search Console screenshots with no commentary is not reporting. Ask what decision the report is meant to help you make.
• Quality control time on a freelancer. Someone still has to check a freelancer's prices, mileage and terminal names against your actual fleet and rates. That's the same 15–20 minutes per article you'd spend on an AI draft, but it's rarely budgeted for either.
• Editing time on AI drafts. Budget 15–20 minutes per article specifically for prices, local road and postcode detail, and anything about your specific vehicles. Skipping this step is how AI content ends up publishing a wrong Heathrow-to-Oxford fare that a customer holds you to.

What a good agency does that no tool does

I'd be doing operators a disservice if I let this read as tools-beat-agencies, because that isn't true for every job. A good agency earns its retainer on work that has nothing to do with publishing volume:

• Technical migrations. Moving CMS platforms, fixing a botched redesign that deindexed half the site, sorting out redirect chains after a rebrand. This is specialist, judgement-heavy work that a content tool has no view of at all.
• PR and genuine link outreach. Getting covered by a regional business paper, getting listed by a wedding venue's preferred supplier page, building relationships with corporate travel booking platforms. This is relationship work done by a human who picks up the phone, and it's worth more per link than almost anything an automated exchange produces.
• Strategy in genuinely competitive markets. In London, Manchester or a market where three well-funded competitors already run the exact playbook in this article, you need someone who can see two moves ahead, not just more pages. That's what a competitive city market for airport transfers looks like once the easy wins are already taken by someone else.
• Accountability with a name attached. When something breaks — a Google update tanks traffic, a competitor outranks you overnight — a retainer means someone is contractually obliged to explain why and fix it. A tool has no such obligation; it's software, and you are still the one accountable for what it publishes under your name.

The honest read is that a tool is very good at the volume-and-consistency problem, which is the problem most 6-car operators actually have, and weak at the judgement-and-relationships problem, which is the problem a handful of operators in saturated markets have instead. Knowing which one you have saves you from buying the wrong fix.

A decision framework by fleet size and market

• 1–4 cars, low-competition town or suburb: DIY with a tool for volume is almost always the right call. You cannot justify an £800+/month retainer on a £150,000–£250,000 turnover, and the market isn't competitive enough to need agency-grade strategy.
• 5–10 cars, moderate competition (most home counties and regional UK markets): This is the model above. A tool covers the publishing volume; spend the saved budget on a one-off technical audit from a freelancer or small agency every 12 months, and on Google Business Profile and review generation, covered in the profile guide for operators.
• 10+ cars, competitive metro market (London, Manchester, Dubai, Sydney CBD): A retainer earns its keep here, specifically for outreach and technical strategy, run alongside a tool for the content volume rather than instead of it. Ask the agency directly whether they'll manage a tool's output as part of the retainer; a good one will say yes without flinching.
• Any fleet size mid-migration or post-rebrand: Pay for agency or freelance technical help first. No amount of new content fixes a site Google can't crawl properly.

The honest conclusion, after building this model for a dozen accounts with slightly different numbers plugged in, is that price alone rarely predicts outcome. What predicts outcome is whether the pages being produced answer a specific search a real customer types, at whatever cost you choose to produce them. A £1,100 retainer that writes six sharp route pages a month will outperform a $99 tool churning out 30 generic ones, and the reverse is just as true. Get the brief right first, then pick the cheapest option that can execute it consistently for twelve months without you having to manage it weekly.