AI applied to marketing is sold as an efficiency lever. But there's a cost almost nobody factors in: constant rework, model maintenance, and human oversight that never fully goes away. These are the same mistakes made with programmatic advertising years ago, when the promise was to cut costs and automate everything.
If you manage ad budgets at an SME or mid-market company, you need to answer one concrete question: do the savings AI claims to deliver account for these hidden costs? If the answer is no, your real ROI is far lower than what the reports show.
This article looks at the parallel between the promise of programmatic and that of AI, and suggests how to evaluate an investment honestly before putting it in your budget.
The Broken Promise of Programmatic: A Costly Déjà Vu
Programmatic advertising, the automated buying and selling of ad space in real time, hit the market with a clear promise: less waste, better targeting, operating costs in free fall. Companies invested. Then they realized that running the platforms required specialized skills, that data had to be cleaned and kept up to date, that ad fraud was eating into budgets, and that creative control had been lost along the way.
Today the same story is repeating itself with AI applied to marketing. Search Engine Journal points it out: before claiming efficiency gains, you need to count the hidden costs of rework and ongoing maintenance. If you don't, you risk making the same mistake again and weakening your long-term strategy.
Where the Costs Really Hide
When a company adopts an AI tool for marketing, whether to generate copy, optimize campaigns or manage attribution, the visible cost is the license or subscription. The invisible one is everything else:
- Rework: AI output has to be reviewed, corrected and adapted to the brand's tone. Who does it, how long does it take, and what does that hour cost?
- Prompt and model maintenance: the instructions you give the AI go stale. Products change, the market changes, platforms change. Someone has to update everything.
- Human oversight: automation doesn't remove accountability. An AI-generated error on a live campaign can cost far more than the savings you gained.
- Integration with existing systems: CRM, ad platforms, analytics. They rarely connect without friction.
These costs don't show up in demos. They show up in the invoices of the following months and in the time taken away from your team.
The Parallel With Campaign Automation Today
This isn't a theoretical problem. Anyone running campaigns on Google or Meta already knows automation is never truly autonomous. We covered this when looking at how to stay in control when Google and Meta Ads automate decisions: platforms optimize for their own goals, not necessarily the advertiser's. Governance remains human work.
The same goes for attribution. Models that distribute credit for conversions across channels can be misleading. We examined the topic in detail in our piece on advertising attribution and the risks to SME budgets: a wrong model shifts budget toward channels that seem to perform, not the ones that actually do.
Adding AI to an already imperfect attribution system doesn't solve the problem. It just hides it better.
How to Assess Real ROI Before You Sign
Before putting any AI marketing solution in your budget, it pays to run an honest calculation. The platform cost alone isn't enough.
- Monthly oversight hours: estimate how many hours your team will spend checking, correcting and updating outputs. Multiply by the real hourly cost.
- Cost of rework: how many times a month is AI-generated content or creative rejected or rewritten? Every rejection has a cost.
- Initial integration cost: development, setup, team training. It's usually underestimated by 40-60%.
- Cost of error: what happens if the tool produces a wrong output on a live campaign? Do you have a control plan?
Only after estimating these numbers does it make sense to compare them with the savings the vendor claims.
What to Do in the Next Few Days If You're Weighing an AI Investment
If you've already adopted AI tools for marketing, or are about to, there are a few concrete actions to put on your agenda this week.
First, ask your team how many hours a month they already spend on oversight and rework of AI output. If nobody knows, that's the first problem to solve: you can't measure what you don't track.
Next, map the integration points: where the AI tool connects with your ad platforms, your CRM, your analytics. Every connection is a potential point of friction.
Finally, consider non-human traffic. Anyone running performance campaigns already knows that part of the budget gets eaten by bots and automated traffic. Adding AI to the mix without a protection system can amplify the problem: we wrote about it in how to protect your ads budget from bots and AI traffic in 2026.
If you want a broader picture of current trends in paid media, start with the advertising, paid media and video section of our blog, where we collect the most relevant analyses for the Italian market.
AI in marketing isn't a scam. It's a tool that works when you adopt it with calibrated expectations and a complete cost calculation. The problem isn't the technology: it's the tendency to buy the promise instead of checking the numbers. Programmatic already taught us this lesson once.
Related articles
Discover more articles exploring similar topics, selected to offer you a more complete and stimulating perspective. Each piece of content is carefully chosen to enrich your experience.
