Five minutes was not long. This is the rest.
We are Analogiq, the AI and martech consultancy you met at the Savoy. We help marketing teams get measurable results from AI, and we are as likely to tell you what not to build as what to build.
The audit is two hours of your time in total, a 30-minute call and a 90-minute session, and ends with one page giving our view on where AI would be worth something in your team and which idea to look at first. If the answer is nothing yet, you will hear that.
It starts with a 30-minute call with Steve and Mario about what you are being asked to do and what you would want from it. That call is what the calendar books.
Rather ask something first? Email, call or send a note.
What does the audit involve?
AI can be applied to almost anything, which is the problem. Most of what it could do in a marketing team is not worth doing. The audit is for pinpointing where it would be worth something in yours, and which idea to look at first.
1. A 30-minute call
What you are being asked to do with AI, what you would want from an audit, and who should be in the session. If we are not the right people, we say so here and that is the end of it.
2. A 90-minute session
In your office or on a call. We list every candidate: the ideas you are being asked about, the ones shelved before, and the parts of the operation where the time and money go. Then we sort them by what each is worth against what it would need, and agree the one to look at first.
3. A day of ours
What we heard, what we think, and where the value is most likely to be. Nothing technical; no systems are examined at this stage.
4. One page, within a week
Sent to you and whoever you ask us to copy. No deck, no meeting required to receive it.
What the page says
The candidates we listed and what we think of each. The one we would look at first, and why. The ones to leave alone, and why. The one thing to settle before spending money on any of it.
What it does not include, and what follows
The audit is free and carries no obligation. It does not include a look at your data or systems, effort estimates or a roadmap; those come from an opportunity assessment, a fixed-price piece of work over about three weeks, which we only propose if the page gives you a reason to want it. If the page says nothing is worth doing yet, that is the page.
What did the room say it was struggling with?
Everyone at the Savoy gave the organisers their top three challenges beforehand. Read together, they come down to these. Each one has a different answer in every business, so here is the question we would start with rather than the answer.
The most common line on the list. The useful first step is a list of what is actually being asked for. What are the three things you are actually being asked to do?
Most teams do not need a new platform to start. Whether the current one can feed a model depends on what is in it and how it is joined. Which system would the first use case have to read from, and who trusts the data in it?
The unglamorous problem. It is rarely all the data; it is usually two sources for one use case. If you could only fix the data for one thing this year, which thing?
Where the output reaches a customer, the question is not whether AI can do the job but where the person sits in the process. What is the last thing a human checks before it goes out, and would that still be true with a model in the chain?
You can only personalise for the people you can recognise. The technology question comes after the coverage question. How much of your audience arrives as a known customer, and what is the first useful message to them?
Activity that touches a lead, a booking or a sale can be measured against a control. Brand cannot be measured the same way and should not be forced to. Which of your current activity has a control, and which never did?
Who we have done this for










Not ready for two hours?
In the week after the Savoy we are writing up what the room said it was struggling with and what we think about it. If you would like it, email either of us and it comes to you when it is done. And if you would rather hear the AI conversation in a room with other marketers, AI in the Wild is the meetup we run in London.