Most AI investment in marketing stalls because tools are added to fragmented data and processes. Deeper commercial value comes from redesigning the marketing operating model, then embedding AI where it improves speed, quality and customer relationships.
AI entered marketing on a promise of speed: faster content, faster insight, faster personalisation. Many marketing leaders have banked early gains but are now confronting harder challenges.
Why are companies using broadly similar AI tools seeing such different results? What enables some teams to shorten campaign cycles while others battle delays, rework and endless handoffs? And perhaps most importantly, how can AI move metrics like conversion, retention and customer lifetime value, rather than just making individual marketers more productive?
From our experience, the answers have far less to do with technology than with the marketing operating model surrounding them. The organisations that are pulling ahead treat AI implementation as a design decision, embedding it at the points in the campaign lifecycle where it improves speed, quality and commercial return.
Why AI investment in marketing stalls
80% of organisations say AI is now a core part of their marketing workflows, according to HubSpot. However, adoption and value are proving to be two very different things, and boards are starting to ask why.
The same three underlying issues appear in almost every marketing function we spend time with:
1. AI is being layered onto slow or fragmented processes
Marketing teams start using AI tools to produce faster copy, image generation and insight summaries. But if the campaign brief is weak, or approvals bottleneck at every stage, then AI is only speeding up one link in a broken chain.
2. Marketing teams aren’t set up to work in an AI-enabled way
Working with AI demands different skills — prompting, editing, testing and working at pace — and in many marketing functions, roles and skills haven’t caught up. Teams have been handed tools without the training or, crucially, the permission to use them properly. AI remains an add-on rather than becoming part of how planners, creatives, media and insight teams collaborate. And, too often, it’s confined to a specialist or innovation team rather than the wider function.
3. Content, data and measurement foundations aren’t strong enough
So many of the big AI opportunities in marketing rest on good inputs: brand rules codified so AI tools can follow them, content assets structured for reuse, centralised customer data, and success metrics aligned across the business.
Fragmented customer data highlights this issue most prominently. We’ve watched brands invest in AI-driven personalisation tools while their transactional data is scattered across channels, and the results can do active damage. If the AI tool can’t see that someone bought a product on Tuesday, it will merrily chase them round the internet with ads for that same product on Thursday. Sometimes with a discount attached.
When Procter & Gamble struggled to personalise its campaigns at scale, it realised the root cause was stale customer insight, drawn from surveys and retail data. Rather than using AI to analyse that data faster, the company redesigned how insight was generated, using connected products like Oral-B toothbrushes to give them a live, personalised view of customer behaviour. The result was live brushing data enabled personalised. Brush head refill notifications delivered a 30% click-through rate and a four-fold increase in repurchase frequency, replacing interruptive, one-size-fits-all advertising with a relationship-based marketing model.¹
Are your marketing processes ready for AI? A simple litmus test
How well are your marketing processes working today? If the answer is “reasonably well”, AI can often make them faster, sharper and more consistent. But if there are fundamental problems, AI investment is unlikely to deliver the results you want.
Take campaign briefing: a pain point we see across global marketing organisations. Often, briefs aren’t integrated across media, creative and customer teams. Objectives are broad, or too vague, or both. And requirements change halfway through at the point teams start collaborating.
There are now some impressive AI briefing tools on the market, trained on the behaviours of top strategists and historic successful marketing briefs. They flag weak points, challenge woolly wording, sharpen objectives, and suggest targeted improvements; well above anything you’d get from a generic large language model (LLM).
But drop those tools into a broken briefing process and their value is constrained. While they improve the document itself, the process, speed and creative thinking remain untouched.
Fixing the process first allows AI to create real commercial value for marketing leaders: not as a sticking plaster for a broken process, but as an accelerator of a better one.
Marketing pain points: process problem or AI opportunity?
One exercise we regularly conduct with leadership teams is separating operational issues from genuine AI opportunities. Getting that distinction right protects both the investment case and the credibility of your wider AI programme.
Most marketing functions will recognise at least one of these pain points:
| If you're experiencing... | Look at the process first when... | AI is likely to add the most value when... |
|---|---|---|
| Campaigns taking too long to launch | Briefs are inconsistent, approvals are duplicated, ownership is unclear and tasks repeatedly come back for rework. | The workflow is well-defined, and AI can accelerate briefing, localisation, versioning, quality assurance or trafficking. |
| Varying content quality | Brand guidance differs between teams, audience needs aren't well understood, or quality standards are inconsistently applied. | The model is working from agreed brand guidance, audience insight and compliance rules. |
| Personalisation not delivering results | Customer data is fragmented, consent is unclear or audience segments are inconsistent. | Tools can use trusted customer data to recommend the next best content, offer or step in the journey. |
| Research isn't influencing campaigns | Insight is difficult to find, scattered across the organisation or disconnected from planning. | AI can retrieve, synthesise and embed insight directly into planning and briefing. |
| Martech not delivering the expected return | Platforms aren't integrated, ownership is unclear or teams lack the capability to use them consistently. | AI tools can automate and optimise activity once integration, ownership and measurement are in place. |
How industry leaders are driving enterprise value from marketing AI
Contained, piecemeal exploration is often how behaviour change starts, and giving people room to test AI tools has its own value. But there is a marked difference between creative exploration and driving enterprise value from marketing AI.
Most organisations are still at the stage of individuals or small teams using AI to move faster. Deeper value can’t be derived from this adoption, no matter how widespread, because the gains are siloed.
Enterprise value comes from redesigning the marketing operating model with AI inside it: the outcomes it should serve, the data it draws on, and the skills and guardrails around it. We see the most successful companies:
- Starting with value, not tools. Be selective about the outcomes you want to improve — speed to market, conversion, production cost, quality of decision-making — and direct AI investment to those specific objectives.
- Sorting the foundations first. Customer data, content assets and brand guidelines need to be centralised, consistent, trustworthy and clearly formatted before LLMs can start using them.
- Treating AI as an operating model decision, not a procurement one. AI should be designed into how campaigns are planned, created, approved, launched and measured, rather than added around the edges.
- Retaining the creative decisions for people. AI is best used for the repetitive aspects of the creative process: pulling insight together, summarising research, creating content variants, adapting assets across channels, and helping teams visualise ideas earlier and quicker.
- Investing in people and behaviour change. Marketers will need training to brief, prompt, review and challenge AI outputs. There are also opportunities to upskill people and reimagine marketing roles with AI handling routine processes.
- Putting guardrails in early. Brand safety, legal review, data privacy and human sign-off should be defined before AI scales. Strong governance will allow your teams to move at speed without compromising consistency or campaign quality.
- Connecting measurement to commercial outcomes. Efficiency gains are the easiest to evidence: Lumen Technologies cut its B2B campaign launch time from 25 days to 9 using AI. But the stronger business cases tie AI to commercial KPIs, like conversion, retention and customer lifetime value.
Above all, industry leaders recognise that marketing is a deeply creative and people-led field, and AI should support the craft rather than substituting for it. For example, Taylors of Harrogate is using AI to enhance the storyboarding process with its creative agencies. Historically a slow, expensive and heavily iterative task, storyboarding has become faster and more accessible with AI image generation, without touching creative ideation.
Every marketer will have AI; not every function will be designed for it
Up to this point, having AI in marketing has been an advantage in itself, but that advantage won’t last. The tools are getting better and easier to access, and within the near term, most marketing functions will be working with similar intelligent capabilities.
Redesigning your operating model for AI — connecting your customer data, codifying your brand and integrating your stakeholders and processes — will ensure technology investments drive acquisition, retention and customer lifetime value.
Your teams will also have more time and headspace for the kind of brave creative and out-of-the-box ideas that make your brand stand out.
Gate One helps marketing leaders design fit-for-future marketing functions, ensuring AI delivers measurable impact. Get in touch, we’d love to talk.
