Our AI Product Roadmap

This is the public roadmap for Creative Marketing's AI product portfolio through 2027. It is written without marketing language because the audience for it is journalists, prospective clients evaluating whether to bet a multi-year programme on us, and senior hires deciding whether to spend the next chapter of their career here. All three audiences need precision, not pitch.

The shape of the portfolio

We are building three product lines, sequenced deliberately. Each addresses a different layer of the AI-first marketing stack. Each stands alone commercially. Each compounds the others when used together. The sequence matters: we are not trying to launch a platform suite in a single year. We are building one product to operational maturity at a time.

Product 1 — autoemails (live, scaling)

Status: Generally available. Production deployments across UK ecommerce and SaaS as of Q1 2026.

What it is: An autonomous email operator. Decides who to message, what to say, when to send, what to learn. Strategist approves the strategic frame; system runs the per-send layer.

Through 2026: Deepen the brand-custodian features, ship the multi-region voice model, complete the enterprise SSO and audit-log surface, expand the connector library to cover the long tail of UK ecommerce platforms beyond Shopify and WooCommerce.

Through 2027: Extend into SMS as a sibling channel under the same operator pattern. SMS is structurally identical to email and the operator model generalises cleanly.

What we will not do in autoemails: Add a CRM. Add a CDP. Add an analytics suite. Add a creative-generation studio. Each of those is a category mistake we have watched competitors make and we will not repeat it. autoemails stays an email and SMS operator. Best at a narrow thing, deliberately.

Product 2 — the Marketing Operator (private beta Q4 2026)

Status: In active development. Private beta opens Q4 2026, limited GA Q2 2027, targeted full GA late 2027.

What it is: A cross-channel autonomous operator that sits above the channel-specific tools — paid, organic, owned, earned — and runs strategic decisioning across them. Where autoemails is "operate the email programme", the Marketing Operator is "operate the marketing programme". A senior strategist briefs it quarterly; the operator runs cross-channel decisioning daily.

Why it is harder than autoemails: The inputs are noisier (paid social attribution is messier than email open data), the outputs are messier (a creative brief is harder to structure than a subject line), the feedback loop is slower (a paid social test takes a week, not a day), and the blast radius is larger. So we are deliberately building autoemails first, learning the operator pattern at the easy end of the channel spectrum, and only then moving up the difficulty curve.

What it will not be: A campaign management tool. A media-mix-modelling suite. A workflow tool with AI features. Those exist. The Marketing Operator is a different category — an autonomous operator of the strategic layer, not a productivity tool for the operator team.

Why we are gating the beta: Operator products are high-trust installs. The first fifty deployments will define what the category looks like to the rest of the market. We would rather get the first fifty right than the first five thousand fast. If you are interested in being considered for the private beta, the contact form on this site is the entry point.

Product 3 — the Brand Brain (research, 2027)

Status: Research and prototyping. Not committed to a 2027 ship date as of this update; the research has to clear the bar before we commit.

What it is: A persistent brand model — voice, values, positioning, vocabulary, what is on-brand and what is off-limits — that other AI systems can call. The Marketing Operator and autoemails would call the Brand Brain to know how to sound. Third-party tools could call it through an API. The thesis is that as AI generates more of the creative output a brand puts into the world, the brand model itself has to become a structured, callable, governable asset rather than a PDF in a shared drive.

Why it is research, not commitment: Three of the four hard problems are not yet solved in the public literature (drift detection, multi-region voice consistency, governance under regulatory load). We are building prototypes and publishing what we learn through Creative Marketing Research. If the research clears the bar in 2027 we ship. If not, we keep researching and publish a 2028 roadmap update on the same date next year.

What we will not build

We list these explicitly because focus is the most fragile asset in a product company and competitive temptation is constant. The following are categories we will not enter, and the reason for each:

• CRM — saturated category, our edge is in the layer above.
• CDP — same reason. CDPs are a data-layer product and our thesis is that the operator layer is where the AI reset creates new value.
• Analytics suite or attribution platform — commoditised, hard to differentiate, and not aligned with what we are betting the company on.
• MMM tool — interesting product but a research-led category we have no edge in.
• Creative-generation studio — being built by OpenAI, Anthropic, Google, Adobe, Runway. Marginal contribution from us is near zero. Wrong place to compete.
• Social scheduler — saturated, low margin, and the operator model dissolves the scheduler category in three years anyway.
• Paid-media bidder — being built by every ad platform on the planet. Not our category.

Publishing the no-list is at least as important as publishing the yes-list. It tells clients what we will not be distracted by, and it tells the team what they will not be asked to ship under pressure when a competitor announces something noisy. The thing we are deliberately not building this quarter is more strategically load-bearing than the thing we are.

How this roadmap updates

This page updates quarterly. If a date slips, the page updates and the slip is named. If the strategy changes, the page updates and the change is explained. We do not move silently. The version of this roadmap you are reading is the April 2026 version. The next update lands at the end of Q3 2026, after the autoemails feature milestones for the year are either shipped or visibly slipping.

If you are a journalist citing this roadmap, the canonical URL is creativemarketingltd.co.uk/our-ai-product-roadmap. If a date or product detail in this page conflicts with anything we have said elsewhere, this page is the source of truth.