
AI-Driven Marketing Automation
Automated journeys across email, SMS and WhatsApp, with a person approving each send.
- Lifecycle journey design & build
- Platform set-up & migration
- Event & trigger infrastructure
What we do · 06 · Marketing Technology
CRM, marketing automation, customer data, lead scoring and measurement that keep marketing working between campaigns.
An illustrative week for one customer: an account is created on Monday and a welcome email is sent; on Wednesday at 22:06 the cart is abandoned and the WhatsApp reminder is held for quiet hours until 09:00 on Thursday; the customer buys on Thursday evening and leaves the journey; a Saturday SMS is not sent because there is no SMS consent.
Why now
Four changes in two years made this true. Each row gives the rule and what it means for your marketing set-up.
01Feb 2024
Mailbox providers
Gmail and Yahoo require bulk senders to authenticate with SPF, DKIM and DMARC, offer one-click unsubscribe (RFC 8058) and keep reported spam below 0.3%.
What it means
Reaching the inbox now depends on sender authentication and complaint rates.
Content & Communication Infrastructure02Mar 2024
Ad measurement
Google’s Consent Mode v2 became the condition for using measurement and personalisation features on traffic from the European Economic Area.
What it means
Tags, consent banners and conversion APIs must agree, or ad platforms optimise on missing data.
AI Campaign Optimization03Aug 2024 → Aug 2026
AI regulation
The EU AI Act entered into force in August 2024; its transparency duties for AI-generated and manipulated content apply from August 2026.
What it means
AI-made creative needs a source record, a label where required and a named approver.
AI Creative Solutions04Nov 2025
Indian data protection
The Digital Personal Data Protection Rules 2025 were notified, bringing the DPDP Act 2023 into operation in phases: notice, specific consent, easy withdrawal and consent managers.
What it means
Consent must be stored per purpose and checked at the moment of sending.
AI-Driven Marketing AutomationDates and rules last reviewed September 2026. Regulations change; we re-check them at the start of every engagement.
How it works
Six illustrative customers enter a journey. Each is checked for consent, holdout, contact limits and approval, and every step is logged. Switch the journey, withdraw consent, approve or reject a send.
Profiles consent per channel
| Profile | SMS | Status | ||
|---|---|---|---|---|
| A-1042New · high intent | Sent | |||
| B-2210Returning | Exited | |||
| C-3307Lapsed · high value | Approval | |||
| D-4415Browsing | Exited | |||
| E-5120Returning | Exited | |||
| F-6093Price-sensitive | Sent |
Approvals 1 waiting
C-3307 · Offers to high-value customers need approval
Next best action: 15% return offer by Email · confidence 0.77
Nothing waiting. Sensitive sends appear here and stop until someone decides.
Audit log newest first
Illustrative profiles and rules. Consent withdrawal is honoured at the next send, as easily as consent was given (DPDP Act 2023 s. 6(4); GDPR Art. 7(3)).
Withdrawals apply at once. A consent withdrawn a minute ago stops the next message.
Every journey has a holdout. One in ten people receives nothing, so each journey’s lift is measured against them.
Sensitive sends wait for a person. Discounts, high-value offers and new AI-drafted copy need a named approver.
7 capabilities
Each is scoped and delivered on its own, and all share one set of customer data, consent records and measurement, so the second capability costs less than the first.

Automated journeys across email, SMS and WhatsApp, with a person approving each send.

Content and messaging tools that route each message to the right channel automatically.

AI proposes targeting, spend and creative changes while a campaign runs, within agreed limits.

AI drafts creative variants from an approved master; a named editor approves each.

Lead scoring and routing that send sales teams the leads worth their time.

Sales follow-ups that adjust to where each prospect is in the buying journey.

Journey mapping, segmentation, loyalty and lifecycle marketing to raise retention and lifetime value.
Five parts, sold and measured as one programme.
Reference architecture
Most martech problems start in a weak layer: a model with no clean data beneath it, or content that cannot reach its channel.
L05Strategy
Who the customer is, which moments matter, what each stage is worth.
L04Channels
Where a decision becomes a message, an ad or a sales action.
L03Content & creative
What is said, assembled from approved facts and components.
L02Decisioning
Journeys, next best action, scoring and budget, each with a fallback.
L01Data, identity & consent
The customer data and consent records every other layer reads.
Events and outcomes flow up from L01. Decisions flow down to L04, checked against consent on the way.
The stack
We work with 45 technologies and choose on fit, total cost and what your team can run. No partner tier is implied.
Where customers are known and reached.
Events, the warehouse and the models that read it.
What is said, sold and shown, from one source.
AI models, data pipelines and the workflows that use them.
How we use AI
Each marketing agent works within written permissions, as you would scope a new hire’s access. The table shows our default; we agree yours with you in the first weeks.
| Agent | Read profiles | Draft content | Build audiences | Send or publish | Move budget |
|---|---|---|---|---|---|
| Brief assemblerPulls goals, segment facts and past results into a campaign brief. | AllowedAggregates | AllowedBrief | Not its job | Never | Never |
| Variant drafterWrites copy and layout variants from approved claims and components. | Not its job | AllowedDraft state | Not its job | Needs approvalFirst send | Never |
| Audience builderTurns a plain-language brief into a segment definition with its size. | AllowedScoped | Not its job | AllowedDraft state | Needs approvalActivate | Never |
| Pre-flight QAChecks links, legal lines, consent scope, rendering and brand rules. | AllowedScoped | NeverFlags only | NeverFlags only | AllowedCan block | Not its job |
| Budget proposerProposes budget moves with the expected effect and confidence. | AllowedAggregates | Not its job | Not its job | Not its job | Needs approvalAbove threshold |
| Lead routerScores and routes inbound leads with the reasons shown. | AllowedScoped | Not its job | Not its job | AllowedApproved templates | Never |
Each agent is tested on your briefs and edge cases for brand, claims and consent, and re-tested after any model or prompt change.
Below a confidence level set for each task, the agent hands over to a fixed rule or a person.
Replies, form fields and web pages an agent reads are treated as data, never as instructions (OWASP LLM01).
Inputs, model and prompt version, output, checks and approver are logged for each step, kept with the campaign and exportable.
Frameworks we build to
We design and document the work against these frameworks, so your compliance and audit teams have evidence to work from. They are frameworks we build to, not certifications we hold.
PCI DSS applies only where quoting and payment flows touch card data. ISO/IEC 42001, the NIST AI RMF and the EU AI Act shape how we document, test and oversee scoring and generative models.
An illustrative consent record for one profile: purpose, channel, notice version and language, how consent was given, where, when, whether it has been withdrawn, and that it is enforced at every send.
Sender standards set up in every build
What changes
We agree five measures before launch and report them in one place. Lift is the difference between people who received a journey and people who did not.
Revenue per customer, journey group minus holdout group
A 10% holdout on every journey · monthly
Active or repeat-purchase rate at 90 days, against holdout
Cohort report · monthly
Unsubscribes and complaints per 1,000 sends; Gmail spam rate kept under 0.3%
Platform and postmaster data · weekly
Incremental cost per acquisition, measured outside ad platforms
Geo experiments and mix model · quarterly
Days from approved brief to a live journey
Delivery log · per release
+12 pts lift at week 12
+7.8 pts lift at week 12
How we work with you
Typically sixteen weeks from audit to an always-on programme. Each phase ends at a gate: a condition that must be met before the next phase starts.
Phase 01Wk 01–03
We review journeys, data sources, consent records, platforms and spend, and measure the gap between platform reports and your own records.
You keep
Gate
Everyone agrees on one conversion definition and the first journey to build.
Phase 02Wk 03–07
Server-side events, identity matching, consent records and sender authentication, built once for every capability that follows.
You keep
Gate
Events arrive in the warehouse and the platform, deduplicated, with consent attached.
Phase 03Wk 06–10
One journey built in full, with its holdout, automatic checks and approval rules, and tested on your customer profiles before it goes live.
You keep
Gate
The journey runs for two full cycles with no consent or deliverability defects.
Phase 04Wk 10–16
The remaining journeys, decision models where there is enough history, agents with written permissions and the measurement set-up.
You keep
Gate
Every journey reports lift against its holdout.
Phase 05Monthly
A monthly review of lift, contact quality and cost. We refit models, plan tests and re-rank the backlog by expected value.
You keep
Gate
Your team can run and change it without us.
What you get
What we make for you is yours once it is paid for. Tools we already had stay ours, and you get a free, permanent licence to use them. The table samples two hand-overs per capability.
| Deliverable | Format | From |
|---|---|---|
| Journey maps & automation plan | Diagrams · board | 01Automation |
| Configured journeys in your platform | Platform build | 01Automation |
| Content model & type documentation | Schema · docs | 02Infrastructure |
| Configured CMS & asset library | Platform build | 02Infrastructure |
| Measurement plan & metric definitions | Doc · sheet | 03Campaign AI |
| Server-side tagging & conversion APIs | Container · code | 03Campaign AI |
| Creative production pipeline | Workflow · code | 04Creative AI |
| Modular template & component system | Figma · templates | 04Creative AI |
| Ideal customer profile & tiered target list | Doc · sheet | 05Lead Gen |
| Intent tracking & enrichment pipeline | Code · config | 05Lead Gen |
| Sales process map & stage definitions | Diagram · doc | 06Sales |
| Configured CRM & automation rules | Platform build | 06Sales |
| Customer journey maps by segment | Figma · PDF | 07Customer Strategy |
| Segmentation model & definitions | Model · sheet | 07Customer Strategy |
Journeys, models, dashboards and templates live in your platforms, your warehouse and your repositories. Nothing sits in an account we control.
Runbooks, admin guides, metric definitions and model cards, written for the people who will run the work.
Every decision rule, prompt and scoring feature is readable, versioned and can be switched off.

Services & packages
You can start with any of these services, bought as a sprint, a fixed-scope project or ongoing capacity. Choose one and your enquiry reaches the right team with it attached.
How to buy
How we work with you
Ask a quick question, send a project brief or issue a formal RFQ. The lead for the work reads each one in full, and any services already in your brief go with it.
Or book a thirty-minute call01
For a first conversation, a press request or anything that does not need a scope yet.
You get A reply from a named lead
02Recommended
Goals, audiences, a budget band and timing, so our first reply can outline the work.
You get Options and a first scope after one call
03
Your documents, deadlines and the procurement and security rules the work must meet.
You get Receipt confirmed and a named bid lead
How it is priced
Each package shows its pricing model. Work starts once a written scope and quote are agreed.
Opens a project brief with this package chosen.
In your brief
In your brief
In your brief
In your brief
In your brief
In your brief
| Package | Every engagement includes | Best for |
|---|---|---|
| SprintA short, fixed-scope engagement that answers one defined question. |
|
Discovery, a diagnostic, a prototype or a decision you need to make soon |
| ProjectA defined scope, delivered for a fixed price. |
|
Work you can describe up front: an identity, a platform or a set of tools |
| MilestoneA larger build in phases you approve and pay for one at a time. |
|
Programmes too large for one contract, where you want control at each step |
| RetainerReserved monthly capacity to run, improve and extend what we built. |
|
Live brands and products that need a steady team without hiring one |
| EnterpriseA multi-workstream programme with a dedicated team, governance and agreed service levels. |
|
Large organisations running change across markets, portfolios or business units |
| SquadA dedicated team that works inside your stack and sprint schedule. |
|
Teams with a clear roadmap that need more senior people quickly |
Questions
For anything else, book a short call. It is usually quicker than a long email.

Start with data and consent foundations plus one revenue journey. Cart recovery, onboarding and win-back are common first choices. That journey needs the event and consent work every later capability uses, and it can show measured lift within a quarter.
Usually not. Most platforms can do more than teams use, so we assess yours first. We recommend a move only when a hard limit, such as real-time events, channel coverage or cost at your volume, cannot be worked around.
In four places: who enters a journey, the next best action, send time and which variant to use. Each has a rule-based fallback, a confidence threshold and a holdout group, so you can measure its effect and switch it off.
A named person approves every campaign send. Agents draft, assemble and check; the send button stays with your team. Triggered transactional and lifecycle messages are approved once as templates, then monitored.
We build to both laws and document the design for your legal team. Consent is recorded per purpose and channel, as set out above. We do not give legal advice; your counsel decides the lawful basis for each purpose.
All three answer different questions. Attribution shows customer paths and helps day-to-day operations. Experiments measure the lift one channel causes. Mix modelling covers the whole budget, including channels you cannot track. We calibrate the model with experiments and use attribution as a diagnostic.
Yes. We can build the stack, data and measurement while your agency runs media and creative. Roles, access and approval rights are agreed in writing at the start, so nothing falls between teams.
You will speak to a lead who would run the work, and get a straight answer on fit.
Three ways to start
Every engagement starts with a written scope and a quote agreed before work begins.
Choose one of the three ways above