Personalisation is overrated for most businesses and underrated for some. Here is how to tell which side of the line you are on.
Personalisation is worth doing when
You have enough traffic for the data to be meaningful. Recommendation engines need volume. If your site gets 500 visitors a month, the data is too sparse to do anything useful with.
The cost of getting it wrong is low. Showing the wrong product recommendation on a retail site is fine. The user just keeps scrolling. Showing the wrong personalisation on a B2B site can feel intrusive. “We know you visited three times last week” makes some buyers nope out of the conversation.
The customer journey is repeatable. Retail sites where users browse, abandon, return are good candidates. B2B sites where each enquiry is bespoke and the customer journey is one visit are not.
You have the technical infrastructure to do it well. CDP, real-time event tracking, content management that supports variant delivery. If any of this is missing, you can stitch something together with tags and pray, but it will be brittle.
Personalisation is not worth doing when
You are doing it because the marketing automation tool came with the feature. The feature exists. That does not mean it should be used. Most personalisation is half-baked because it was set up to justify the licence cost.
The user does not benefit from it. “Hi Brian, since you are reading our consultancy page, here is our consultancy page” is not personalisation. It is a meta-comment.
The data you would personalise on is unreliable. If your CRM data is 30% wrong, your personalisation is 30% wrong, which means it is mildly creepy 30% of the time.
The honest version: most B2B service businesses do not need personalisation. They need a clear value proposition, an obvious next step, and a contact form that works. The maths on a marketing site that converts at 2% gets fixed by getting it to 4%, not by personalising for the 96% who leave.
If you are debating whether to invest in personalisation, start there.