For years, email marketers have talked about personalisation as though it is automatically a good thing.
Use the customer’s first name. Show products they browsed. Recommend similar items. Segment by purchase history. Mention their location. Change the hero image. Trigger a message based on behaviour.
And yes, all of those tactics can be useful.
But none of them are automatically valuable. Because personalisation is not valuable simply because it uses data. It is valuable when it helps the customer make a better decision, reduces friction, improves relevance, supports timing, or makes the experience feel easier and more useful.
That distinction matters. Especially now. With more data, more automation, more AI-assisted content creation, and more tools promising one-to-one experiences at scale, it is becoming easier than ever to personalise email. But easier personalisation is not the same as better personalisation.
And if marketers are not careful, personalisation can quickly become noise, creepiness, or complexity wearing a clever little name badge.
Personalisation is not the goal
One of the biggest mistakes marketers make is treating personalisation as the objective.
It is not.
The objective is not to prove that we know something about the customer. The objective is to make the communication more useful, relevant, persuasive, or timely because of what we know. That is a very different thing.
A first name in a subject line is not automatically personal. A product recommendation is not automatically relevant. A location-based message is not automatically helpful. A behavioural trigger is not automatically well-timed.
The question should never be, “Can we personalise this?”
The better question is, “Will this personalisation improve the customer’s experience or decision?”
If the answer is no, it may not be personalisation at all. It may just be data decoration.
The difference between data-driven and customer-led
A lot of email personalisation is data-driven. That sounds impressive, but it is not enough.
Data-driven personalisation starts with the data. We know this customer browsed running shoes. We know they bought skincare last month. We know they live in Manchester. We know they clicked on summer dresses. We know they have not purchased for 90 days.
So, we use that data to change the email.
Customer-led personalisation starts somewhere else. It starts with the customer’s likely need, mindset, question, hesitation, or next best step.
If someone browsed running shoes, are they comparing options? Replacing an old pair? Training for something? Looking for comfort? Sensitive to price? Unsure about sizing? Trying to avoid injury?
If someone bought skincare last month, do they need usage guidance, replenishment timing, routine support, complementary products, reassurance, or simply time to experience the product before being sold to again?
If someone has not purchased for 90 days, are they dormant, dissatisfied, distracted, no longer in need, waiting for a discount, buying elsewhere, or simply on a longer buying cycle?
The data may tell us what happened. It does not automatically tell us why it happened. And the why is often where better email marketing begins.
Relevance is not the same as recognition
Sometimes personalisation simply proves that the brand recognises the customer. You looked at this. You bought this. You live here. You clicked that. You have not purchased recently.
That can be useful. But recognition alone is not enough.
In some cases, it can even feel lazy or intrusive. The customer does not necessarily need to be reminded that you have data about them. They need that data to be used in a way that makes their experience better.
There is a big difference between “You viewed this product. Buy it now” and “Still deciding? Here’s how to choose the right size.”
There is a big difference between “You bought this serum. Here are three more products” and “Here’s how to get the best results from your new serum.”
There is a big difference between “You haven’t shopped for a while” and “Still interested in hearing from us? Tell us what would be most useful.”
The first examples are brand-led. They use data to push the next action the business wants.
The second examples are customer-led. They use data to support the customer’s next decision.
That difference is subtle. It is also huge.
Personalisation can reduce friction
Good personalisation often works because it reduces effort.
It helps the customer find the right thing faster. It removes irrelevant choices. It anticipates a question. It recognises where someone is in their journey. It makes the next step feel more obvious.
This is particularly important because customers are not always making slow, careful, fully rational decisions. They are often scanning, comparing, hesitating, multitasking, and trying to work out what matters.
A well-personalised email can make that easier.
A returning customer may not need the same introductory brand story as a new subscriber. A first-time buyer may need reassurance, education, and confidence before they are ready for another purchase. A high-intent browser may need comparison help, reviews, sizing information, or delivery reassurance. A customer who repeatedly buys the same product may value replenishment timing more than a broad product recommendation. A subscriber who only engages with educational content may not be ready for a hard sell, but may be ready for a deeper guide, webinar, or course.
In each case, the value of personalisation is not the technical act of changing the content.
It is the reduction of cognitive effort.
The email feels easier to process because it better matches the customer’s context.
Personalisation can also add friction
Of course, personalisation does not always make things easier.
Sometimes it adds friction.
Too many dynamic modules can make an email feel disjointed. Over-specific targeting can make a customer wonder how much the brand knows. Bad recommendations can make the brand look careless. Over-personalised subject lines can feel awkward. Lifecycle triggers can arrive at the wrong moment. Automated emails can keep nudging someone towards an action they have already taken. And AI can generate endless personalised variations without improving the underlying relevance.
That is the danger of treating personalisation as a technology problem rather than a judgement problem.
More variables do not automatically mean a better experience. More data does not automatically mean more relevance. More automation does not automatically mean better timing.
Sometimes the most customer-led decision is not to personalise more.
Sometimes it is to personalise less, but better.
The creepiness test
A useful question for marketers is this: Would this feel helpful if the customer noticed it?
That is the creepiness test. Some personalisation works best when it feels natural and supportive. The customer may not consciously register every personalised decision, but the experience feels smoother.
Other personalisation becomes uncomfortable the moment the customer notices it.
A reminder based on a product they nearly bought may feel useful if it includes sizing advice, reviews, or availability information. But if it simply says, “We saw you looking at this,” it may feel more like surveillance than service.
A birthday email can feel warm if it is genuinely celebratory or useful. But if it is simply a thinly disguised sales push, the personal touch starts to feel transactional.
A location-based email can feel helpful if it gives genuinely relevant information. But if it uses location simply to appear clever, it adds little value.
The issue is not whether the brand has the data. The issue is whether the customer experiences the use of that data as helpful, respectful, and proportionate.
AI makes judgement even more important
AI is going to make personalisation easier to scale.
It can help generate different versions of copy for different segments. It can summarise customer behaviour. It can suggest product recommendations, lifecycle triggers, content angles, subject lines, and next-best-action ideas.
Used well, this can be incredibly useful.
But AI does not remove the need for human judgement. In fact, it increases it.
Because when it becomes easy to generate personalised variations, marketers need to become much better at deciding which variations should exist.
Just because AI can create ten versions of an email does not mean all ten are strategically necessary. Just because a platform can personalise based on behaviour does not mean the behaviour is meaningful. Just because a system can predict a next-best product does not mean the customer is ready for another product. Just because a trigger can be automated does not mean the timing is right.
AI can help marketers create more personalised content.
But marketers still need to understand whether that personalisation improves the customer experience, supports the customer journey, and contributes to a meaningful business outcome.
Otherwise, we are not creating better email. We are simply creating more elaborate email.
Personalisation needs a purpose
Before personalising an email, marketers should be able to answer a simple question:
What is this personalisation meant to do?
Is it meant to reduce choice? Increase confidence? Remove irrelevant content? Support a repeat behaviour? Help someone compare options? Reassure them after purchase? Encourage product adoption? Recognise loyalty? Avoid sending the wrong message? Move someone to the next stage of the journey?
If there is no clear purpose, the personalisation may not be worth doing.
This is especially important because personalisation often adds operational complexity. It may require cleaner data, better segmentation, dynamic content rules, more QA, more testing, more creative variations, and more reporting.
That complexity can be worthwhile.
But only when it improves the outcome.
Otherwise, the team may spend hours creating a more complex version of an email that does not become more useful to the customer or more valuable to the business.
And that is not sophistication.
That is busywork with merge fields.
How to make personalisation more customer-led
A more useful approach is to begin with the decision the customer is trying to make.
What does the customer need to know? What are they unsure about? What might stop them from acting? What context do we already have? How can we use that context to make the email easier, clearer, or more relevant? What would be inappropriate, excessive, or unnecessary?
This shifts personalisation away from “what data do we have?” and towards “what help can we provide?”
For example, instead of personalising only by product category, personalise by buying stage.
A new subscriber may need orientation. A browser may need comparison support. A first-time buyer may need reassurance. A repeat buyer may need convenience. A lapsed buyer may need renewed relevance. A loyal customer may need recognition.
The same data can be used very differently depending on the customer’s likely mindset.
That is where better personalisation lives. Not in using more data, but in using data with more empathy and purpose.
Different types of personalisation do different jobs
Not all personalisation is the same.
Some personalisation is identity-based. It uses information such as name, location, birthday, company, or stated preferences. Some is behavioural. It responds to what someone has browsed, clicked, purchased, downloaded, abandoned, renewed, or ignored. Some is lifecycle-based. It adapts the message according to where someone is in the customer journey. Some is predictive. It uses modelling or AI to estimate what someone may need, want, or do next. Some is contextual. It uses timing, seasonality, device, weather, stock availability, or external conditions. Some is preference-led. It responds to what the customer has explicitly asked to receive.
Each type can be useful. Each type can also be misused.
A first name can feel friendly or gimmicky. A behavioural trigger can feel helpful or pushy. A predictive recommendation can feel relevant or baffling. A lifecycle message can feel timely or robotic. A preference-led email can feel respectful or ignored if the brand fails to honour it properly.
That is why personalisation cannot simply be judged by whether it exists. It needs to be judged by whether it improves the experience.
Measure personalisation by value, not cleverness
Another trap is measuring personalisation only by short-term interaction. A personalised subject line may increase opens. A dynamic product block may increase clicks. A triggered message may generate revenue.
Those are useful signals. But they do not tell the whole story. Good personalisation should also be judged by whether it improves customer outcomes and long-term business value.
Does it reduce friction? Does it increase confidence? Does it improve product adoption? Does it reduce returns or complaints? Does it encourage a second purchase? Does it increase retention? Does it improve margin, or simply increase discount dependency? Does it make the customer more likely to trust future emails? Does it help the brand send fewer irrelevant messages?
That last point matters. One of the best uses of personalisation is not just deciding what to send. It is deciding what not to send.
Suppressing irrelevant emails, changing cadence, excluding customers from messages they do not need, or adapting content based on where someone is in their journey can be just as valuable as showing a personalised product recommendation.
Sometimes, the most respectful personalisation is silence.
Why this matters for email marketers now
Email personalisation is only going to become easier to execute.
AI, automation, predictive tools, dynamic content, and increasingly integrated platforms will make it possible to create more tailored experiences at greater speed.
But the marketers who succeed will not be the ones who personalise the most. They will be the ones who personalise with the most judgement.
They will understand customer journeys, buying stages, behavioural triggers, persuasion, testing, data quality, privacy expectations, and measurement. They will know when to personalise, when not to, and how to make personalisation feel useful rather than intrusive.
That is why these skills sit at the heart of the Holistic Email Academy. Our strategy courses help marketers connect personalisation to commercial goals and customer needs. Our lifecycle and automation courses help marketers design journeys that respond to customer behaviour and timing. Our psychology and persuasive copywriting courses help marketers understand what customers need to feel, believe, and know before they act. And our testing and optimisation courses help marketers measure whether personalisation is genuinely improving behaviour, not just adding complexity.
Because personalisation is not about proving how much data you have. It is about using what you know to make the customer experience better.
Personalisation is a responsibility
The future of personalisation is not simply more dynamic content, more predictive recommendations, more triggers, or more AI-generated variants.
It is better judgement.
Personalisation should help customers make better decisions. It should reduce friction. It should make emails more relevant, useful, respectful, and timely. It should support the relationship, not just the next transaction.
That means marketers need to ask better questions. Not just, “Can we personalise this?”
But should we? Why? For whom? At what point in the journey? What customer need does it support? What behaviour are we trying to influence? Will the customer experience this as helpful? And would the email still make sense if the personalisation failed?
Because when personalisation is done well, it can make email feel more useful, more relevant, and more human.
When it is done badly, it can make email feel noisy, creepy, or strangely hollow. The difference is not the technology. The difference is the judgement behind it.
