Insight 2027
Your CRM used to decide how well you grew. Now it determines what you can build.
I want to be careful here, because there is an old version of this argument, and it is boring. The old version says a CRM improves follow-up, tidies the pipeline, makes lifecycle marketing possible, and therefore pays for itself. That has been true for twenty years, everyone already agrees with it, and nobody changes their plans because of it.
If you are not already doing this, you should be; it’s the bedrock of marketing and sales success today.
What I am describing is a different thing that happened quietly in the last two years, and our research this year gave me the numbers to say it plainly. The CRM stopped being a tool that made growth more efficient and became the thing that determines what kinds of growth a business can achieve at all. Not how fast. Whether.
AI did not democratise growth. It put a door on it.
The consensus position is that AI levels the field, because a two-person business and a two-thousand-person business can buy the same model. Our data says the opposite, and the mechanism is specific enough to plan around.
Experimentation with AI is universal. Every kind of business is trying things, at roughly the same rate, regardless of how well equipped it is. Deployment is a completely different story, and it divides along one line.
Sort the AI functions by how close they sit to revenue, and the split is obvious. The functions that need nothing but a subscription- writing, design, general assistance- are running everywhere. The functions that touch revenue- personalisation, lead scoring, sales enablement, customer insight- are running roughly four times as often in businesses with a working CRM as in businesses without one. Seven in ten businesses with no CRM have none of them in production. And the gap on the subscription tier is about half as wide, which tells you it is not about budget, sophistication, or appetite. It is about whether there is any customer data for the thing to point at.
So the tooling has quietly separated into two tiers. One tier anyone can buy produces assets faster and cheaper, and your competitors bought it the same week you did. And there is a tier that only works if the business already knows who its customers are, what they cost and what they are worth. The first tier lowers your cost per asset. The second is where revenue moves. The CRM is the border between them.
That is the shift. Nobody is locked out of AI. Plenty of businesses are locked out of the half of it that pays.
Your customer data stopped being an asset. It became a licence.
An asset sits on a balance sheet and appreciates quietly whether you use it or not. A licence is different. A licence is what you must hold to operate, and it comes with conditions.
The condition on customer data is not technical. This is the part most businesses get wrong, and it costs them years.
A CRM only works when a business has agreed what a customer is, when an enquiry becomes one, who is accountable at each stage, and what counts as success. That agreement is between marketing and sales, and in my experience most organisations have never made it. They both have versions, the versions disagree, and nobody has ever had to reconcile them because no single system has ever forced the question.
Buy the software without agreeing, and you have bought a filing cabinet and two conflicting reports. The evidence for that is the sharpest thing in our study this year. Businesses running a CRM properly grew their marketing budgets around half as often as businesses without one. But businesses that owned a CRM that wasn’t working grew less often than businesses that had never bought one at all, and they were the group most likely to cut their marketing spend. Half of them have no revenue-facing AI in production. Owning the tool and not running it is not neutral. It is a worse position than not owning it, because it comes with the cost, the false confidence, and the conclusion at the end that the software was the problem.
This is now the market’s dominant failure mode, and it is not a buying failure. Across every category of marketing technology, roughly two in five tools that businesses own are described by their own owners as idle or ineffective. Of the businesses that do own a CRM, fewer than one in five uses it to see where a customer sits in their own lifecycle. The market has finished buying. It has not started agreeing.
Waiting has stopped being free.
Every previous version of this decision could be deferred. You could run on spreadsheets and instinct for five years, buy a CRM in year six, import what you had and be roughly where you would have been. The cost of delay was flat.
That is no longer the case, and this is the part I would put in front of a board.
Capability built on customer data compounds on the history you have collected. Personalisation, scoring, forecasting, retention modelling: none of it works on a dataset that starts today. So a business beginning its data layer in 2028 is not twelve months behind its competitor. It is twelve months of customers behind, and that distance keeps opening while both of them stand still. The gap is no longer a gap. It is an interest rate.
And the market is not paying it down. When we looked at what businesses intend to invest in over the next two years, intent followed existing capability rather than compensating for its absence. Businesses with no CRM were more than twice as likely as businesses running one to have no plan at all to build revenue operations, marketing operations or conversion capability within two years. The training priorities across the market are led by social media and content strategy, in a market where almost everyone already runs social. We are collectively planning to get better at what we are already best at.
Meanwhile, the money is moving toward businesses that can account for it. Marketing budgets in South Africa are not being cut, which surprised me given the mood of the past two years, but where they grew tracks data maturity far more closely than size or sector. Businesses at the top of the maturity ladder grew their marketing budgets at more than double the rate of those with no data strategy. Nobody is being denied budget for having bad ideas. They are being held flat because they can’t show what happened in the last round.
The growth is in the base, and the base is in the instrument
Here is why this is a growth argument and not an operations one.
South African marketing measures acquisition roughly twice as well as it measures anything that happens after a customer arrives. Retention, lifetime value, lifecycle progression, satisfaction: four in ten businesses track none of them. The technology category whose entire purpose is adding value to existing customers is the least installed and least functional part of the average stack.
Now set that beside the best-performing content asset in our whole study. Rated above everything a marketing team writes about itself: customer evidence. Case studies, testimonials, proof. Your existing customers doing your selling. Almost nobody manufactures it systematically, because doing so requires knowing which customers are happy, which is a CRM question.
So the cheapest growth available in 2027 is not in reach or in new channels. It is in customers a business has already paid to acquire, in expansion, retention and advocacy it currently cannot see. That growth does not need a bigger budget. It needs an instrument. And it happens to live in the same place that wins the budget argument, which is the closest thing to a free lunch in any of this.
What I would do with the year
I would not buy anything.
I would take the instrument the business already owns and most needs to trust, which for nearly everyone is the CRM, and I would get marketing and sales into a room to agree on what a customer is before anyone touches a setting. That conversation is uncomfortable, takes a fortnight, and is the actual work. Then I would instrument what happens after the sale, because that is where the revenue nobody is counting sits. Then, last, I would point AI at it, and it would work, because by then there would be something for it to read.
The businesses that pull away next year will not be the ones with the best ideas or the biggest budgets. They will be the ones that spent 2027 building, while their competitors spent it comparing tools.
What I would do with the year
Mike Saunders is CEO of Digitlab.
