Marketing automation consultant: what it costs and what it fixes

A marketing automation consultant maps how an enquiry travels through your business, finds the points where it stalls, and automates those. In a small business that is usually follow up and handoffs rather than campaigns. Expect £250 to £500 a day in the UK, or one to four thousand pounds for a first project.

The harder question is whether the thing costing you money is your marketing at all.

Who the other guides are written for

I read the two articles ranking above this one before writing it. Both are excellent and neither is written for you.

They assume a marketing team. They assume a Marketo or Pardot licence, a MarTech stack, and somebody whose whole job is marketing operations. The prices quoted are $175 to $375 an hour.

That is a real job and those are real consultants. It is just a completely different business from a twelve person firm in Timperley wondering why enquiries go quiet.

This is the version for one to fifty people.

The two per cent problem

The thing I say most often in training rooms is that people use these tools at about two per cent of what they can do.

Not because they lack skill. Because they have not yet believed the thing could handle a bigger job, so they keep giving it small ones.

Somebody uses it to tidy up an email. That works, so they use it to tidy up the next email. Six months later they are still tidying up emails and have concluded that AI is quite good at tidying up emails.

Marketing automation goes wrong in the same shape. Somebody buys a tool that could run a whole process and uses it to send one newsletter.

Where the money actually leaks

The bit that costs a small business money is almost never writing the email.

It is the handoffs. An enquiry lands in the website inbox. Somebody reads it, decides it is real, copies the name and number into the CRM, tags it, remembers to follow up on Thursday, forgets, remembers the following Tuesday.

Nobody describes that as a marketing problem. It gets described as being busy.

Before you talk to anybody about automation, take one enquiry from the moment it arrives to the moment a human has actually spoken to that person, and write down every point where somebody moves information from one place to another.

Most small businesses I look at have between four and seven. Each one is a place an enquiry can quietly die.

That exercise is free, takes half an hour, and is a better brief than most consultants get given.

What one actually does

Four things, once you strip out the language.

They map that route properly, including the parts that happen in somebody's head or on a sticky note.

They decide what is worth automating. Some of it is. A fair amount is not, because it happens rarely or because the judgement matters more than the speed.

They pick tools, and for a small business that usually means using what you already pay for rather than buying a platform.

Then they build it and show somebody how to run it. That last part is the one people skip, and it is why automations quietly stop working about four months in.

The enterprise ideas, translated

The guides on this term are full of concepts that are genuinely useful in a large company and slightly daft at twelve people. Here is the small business version of each.

Lead scoring at enterprise means a model assigning points to behaviours. At your size it means deciding, out loud, which enquiries you ring back first. Write the rule down. That is the whole thing.

Nurture campaigns means a sequence of emails over months. At your size it usually means one email, sent three days later, to the people who asked for a price and went quiet. That single email outperforms most sequences because almost nobody sends it.

Sales and marketing alignment means service level agreements between departments. You do not have departments. It means deciding who rings the enquiry and by when.

MarTech stack means an integrated platform architecture. It means your inbox, your calendar and whatever you keep customer details in, talking to each other well enough that nobody retypes anything.

If a consultant uses the enterprise vocabulary without translating it, they are describing work they have done somewhere else.

Which tools, honestly

Almost certainly not HubSpot's paid tiers, and definitely not Marketo or Pardot.

Those platforms are priced and designed for teams with somebody running them full time. Bought by a small business, they become an expensive place to store contacts, which is a thing a spreadsheet already does for nothing.

The Department for Science, Innovation and Technology asked UK businesses what held them back on AI. Seventy one per cent said they had not identified a business need. Sixty per cent cited limited skills. Cost came further down.

Which matches what I see. The tool is rarely the constraint. Nobody has decided what the tool is for.

For most firms of your size, the answer is the CRM you already have, a scheduling link, and one alert that tells somebody an enquiry has been sitting unanswered. That is unglamorous and it is where the money is.

What it costs in the UK

Independent consultants doing this work with small businesses sit somewhere between £250 and £500 a day. I am at that end of the market rather than the agency end.

A first project, properly scoped, is a few days rather than a few weeks. One to four thousand pounds gets most small firms a working set of automations and the training to run them.

A diagnostic conversation should be free. Nobody should charge you to work out whether they can help.

If the number quoted has a comma in it and the work has not been broken into pieces you can buy one at a time, be careful. The big number gets deferred. The small one gets started.

Be careful with the time savings you are quoted

Including mine.

Research by Opinium for OpenAI found UK small businesses using AI report saving 5.2 hours a week. That was funded by an AI company and it is self reported.

The Federal Reserve Bank of St Louis, measuring rather than asking, put it nearer 2.2 per cent of working hours. That is about an hour.

Those two numbers describe the same activity and differ by roughly a factor of five. I would work from the smaller one. It is duller and more likely to be true, and a consultant quoting you the bigger one is telling you something about the consultant.

An hour a week is still fifty hours a year. And that average includes everybody using it badly.

Why the skill matters more than the software

Anthropic published something from its own usage data, with the obvious caveat that it is describing customers of its own product. People using these tools for six months or more were a few percentage points more likely to have a productive session, and were putting in noticeably more complex requests.

More complex requests. That is the whole thing.

The tool did not improve. The person got braver.

Which is why a consultant who configures your software and leaves has sold you the smaller half of the job. The software will do what it is told. Whether anybody in your business ever tells it to do something ambitious is a different question, and it decides whether you got value.

When to hire somebody, and when not to

Hire when you have identified the jobs and cannot get them done, or when you have tried and it half works and nobody trusts it.

Do it yourself when the job is small and somebody in the business enjoys this sort of thing. That person exists in a surprising number of firms and is usually doing something else.

Do neither if the real problem is that you are not getting enough enquiries. Automation makes an existing process faster. It does not create demand, and a faster route to nowhere is still nowhere.

I have talked people out of this work on exactly that basis, and it is the most useful conversation in the diagnostic.

How to tell whether it actually worked

This is the part almost nobody writes about, including the guides ranking above this one, and it is the part I find hardest to do honestly.

The problem is that time saved is very easy to claim and very hard to see. An automation takes forty minutes off somebody's week. Where did the forty minutes go? Usually into other work, which is fine, but it means nothing on a report and nobody can point at it afterwards.

So measure the thing at the edge of the business instead, where it either happened or it did not.

Response time is the best one for marketing automation. Before you change anything, take your last twenty enquiries and work out the average gap between arriving and being spoken to. Write the number down. Do the same thing again six weeks later on the next twenty.

That comparison is crude and it is not a controlled experiment. It is also considerably better than nothing, and it is a number you can get without buying any software.

Conversion is the other one, and it is slower. What proportion of enquiries turned into a quote, and what proportion of quotes turned into work?

Those move for plenty of reasons that have nothing to do with automation, so do not credit a change to the software and stop thinking about it. Seasonality, one big customer, a competitor going quiet.

But a follow up process that went from patchy to consistent ought to show up there eventually. If after six months it has not, the assumption was wrong rather than the build, and that is worth knowing too.

The thing not to measure is emails sent. Every platform will report it enthusiastically because it is the number that always goes up.

I have had this go against me. A workflow I built for a client saved real time and made no difference to anything anybody could see, because the constraint was somewhere else entirely. That was worth finding out, and it was worth finding out in six weeks rather than a year.

What good looks like after three months

Modest, is the honest answer.

Two or three automations running that nobody thinks about. An alert that catches the enquiry nobody chased. A follow up that goes out whether or not anybody remembered. Perhaps a report that builds itself on a Monday morning instead of somebody clicking through a system for half an hour.

Nobody in the business describing themselves as doing marketing automation. It just being how the work happens now.

If after three months you have a platform, a login nobody uses, and a consultant on a retainer, something went wrong at the first conversation.

What you could do on Monday

Take your last twenty enquiries. Not your best ones. The last twenty.

For each, write down the date it arrived and the date somebody actually spoke to that person.

If the average gap is more than a day, you do not have a marketing problem and more automation will not fix it. You have a follow up problem, and the first thing worth building is the boring alert that tells you an enquiry has been sitting for eighteen hours.

Nobody wants to buy that project. It is usually the one that pays for itself first.

A word on choosing a platform, if you must

If you have decided you want a proper marketing tool rather than stitching together what you have, three questions save most of the regret.

Does somebody in the business actually want to run it. Not tolerate it. Want it. A platform without an owner becomes an expensive contact list within about two months, and I have watched that happen more than once.

Can you get your data back out. Ask before you sign, not when you are leaving. Some of the cheaper tools make export deliberately awkward, which you only discover at the point you have three years of history in there.

And what does it cost at three times your current contact count. Nearly all of these price on volume, and the number quoted is for where you are now, not where you say you are going.

None of that is about features. Features are rarely the thing that goes wrong.

The honest limit

Automation makes a good process faster and a bad process worse, at speed.

If your follow up is inconsistent because nobody has decided who owns it, automating it produces consistent, prompt, well written messages from a business that still does not know who is picking the phone up. You will have industrialised the confusion.

Decide who owns the enquiry first. Then automate. That order round is not negotiable and it is free.

And be wary, including with me, of anyone who arrives with a platform in mind before they have seen how your enquiries actually move. The tool should be the last decision, not the first.