Personalisation technology can give eCommerce teams access to increasingly sophisticated capabilities. The harder question comes after implementation: What should actually be personalised?
A retailer could theoretically change website content, recommendations, emails, offers and customer journeys for hundreds of different audiences. Doing so does not automatically create a better shopping experience.
Good personalisation should make the journey more relevant, easier or more useful.
That might mean helping a returning shopper rediscover products they viewed previously, stopping acquisition messages once somebody has purchased, surfacing complementary products at the right moment or providing an incentive only when it is genuinely needed.
Poor personalisation does the opposite. It adds noise, repeats information customers already know or reveals so much behavioural knowledge that the experience becomes uncomfortable.
For eCommerce teams, successful deployment therefore starts with customer problems and commercial objectives rather than personalisation features.
From Personalisation Technology to Personalisation Strategy
Our earlier guide to eCommerce personalisation looked at what buyers should compare when selecting platforms and specialist partners.
Once the technology is in place, the operational challenge changes.
Teams need to determine:
- Which customer journeys to personalise
- Which data to use
- Which triggers indicate genuine intent
- Which products or messages to surface
- Which channels should work together
- Where automation needs human control
- How performance will be measured
A useful starting point is to avoid trying to personalise everything simultaneously.
Instead, select a small number of customer problems where greater relevance could make a measurable difference.
For example:
Problem: Visitors struggle to discover relevant products.
Personalisation: Behaviour-based recommendations.
Or:
Problem: Customers browse repeatedly without purchasing.
Personalisation: Timely browse-abandonment communication.
Or:
Problem: Discounts are being offered too broadly.
Personalisation: Target incentives according to customer intent.
This creates a much clearer relationship between personalisation and commercial outcome.
Start with Customer Behaviour
Behavioural information is one of the strongest foundations for eCommerce personalisation because it shows what customers are actually doing.
Useful signals might include:
- Products viewed
- Categories browsed
- Searches
- Time on site
- Basket additions
- Checkout activity
- Purchases
- Previous visits
- Email interaction
Consider three customers.
Customer A visits a fashion retailer for the first time and looks at coats.
Customer B has returned three times and repeatedly viewed one specific coat.
Customer C added that coat to their basket but did not complete checkout.
All three are technically interested in the same product.
But their behaviour suggests very different levels of intent.
An effective personalisation strategy should recognise those differences.
Dotdigital, for example, currently supports behavioural targeting and automated product recommendations based on signals including product views, cart behaviour and purchase history.
Dotdigital – Personalisation – https://dotdigital.com/personalization/
Build Personalisation Around the Customer Journey
Rather than developing isolated campaigns, retailers can map personalisation across stages of the customer relationship.
Discovery
Help shoppers find relevant categories and products.
Consideration
Use browsing behaviour, reviews and recommendations to support evaluation.
Basket
Recommend complementary products or remove unnecessary distractions.
Checkout
Keep the experience focused and recognise known customers.
Post-purchase
Stop promoting products already purchased and provide useful follow-up information.
Retention
Use purchase history and preferences to support replenishment, cross-selling or re-engagement.
The important point is that personalisation should reflect where the customer currently is, rather than relying on one static customer profile.
Dotdigital describes modern personalisation in similar lifecycle terms, differentiating first-time customers, returning visitors and cart abandoners according to behavioural and contextual signals.
Dotdigital – Complete Guide to eCommerce Personalisation – https://dotdigital.com/blog/the-complete-guide-to-ecommerce-personalization/
Use Product Recommendations with Purpose
Product recommendations are one of the most visible personalisation techniques.
But there is a considerable difference between displaying:
“Popular products”
and:
“Products relevant to this individual customer’s current behaviour.”
Recommendation strategies can include:
- Recently viewed
- Similar products
- Frequently purchased together
- Complementary products
- Replenishment
- Best sellers within a preferred category
- Products associated with current basket contents
Dotdigital’s current recommendation tools can use behaviour and context to support discovery, cross-selling and upselling.
Dotdigital – Recommendation SmartBlocks – https://personalization.help.dotdigital.com/en/articles/8523414-recommendation-smartblocks
But automated recommendations need sensible commercial constraints.
Retailers may need to suppress:
- Out-of-stock products
- Products already purchased
- Incompatible products
- Low-margin products
- Items unavailable in the customer’s region
Personalisation Principle
An algorithm should choose from products the business is actually able and willing to sell.
Automation and merchandising judgement need to work together.
Stop Recommending What Customers Have Already Bought
One of the simplest ways personalisation can fail is by ignoring completed transactions.
A customer:
views shoes → buys shoes → receives emails recommending exactly the same shoes for two weeks
This demonstrates that different parts of the technology stack are not sharing information quickly enough.
Instead, a purchase should potentially trigger a change in journey:
Shoes purchased → suppress shoe acquisition → recommend care product or complementary accessory
This sounds straightforward but depends on:
- Accurate purchase data
- Product taxonomy
- Timely synchronisation
- Appropriate campaign rules
Personalisation therefore becomes partly an integration problem.
Connect Customer and Product Data
Effective eCommerce personalisation sits between two important datasets:
Who is the customer?
and
What can we sell them?
Customer information might include:
- Behaviour
- Purchase history
- Loyalty status
- Preferences
- Location
Product information might include:
- Category
- Brand
- Price
- Availability
- Margin
- Attributes
- Compatibility
The strongest personalisation occurs when those datasets interact.
For example:
Customer frequently browses Brand A + Brand A product available in preferred size → recommendation
rather than simply:
Customer browsed clothing → recommend clothing
Agencies and technology partners such as Magebit, On Tap, CTI Digital, Sozo Design, Arsenalia and VINSINFO can play an important role here because personalisation often depends on correctly integrating eCommerce platforms, customer data and supporting systems.
Personalise Across Channels – But Keep Them Connected
Customers do not experience a retailer as separate technology platforms.
They may:
browse the website → receive an email → return on mobile → contact customer service → purchase
Personalisation should ideally reflect what has already happened elsewhere.
Otherwise:
Website: Welcome, new customer!
Email: Thanks for your third purchase.
The customer notices the inconsistency even if the technology team understands why it happened.
Platforms such as Dotdigital and Brevo support customer segmentation and automated communications across multiple channels.
Brevo – Marketing Platform – https://www.brevo.com/products/marketing-platform/
The goal is not necessarily to personalise every possible channel.
It is to ensure the channels being personalised do not contradict one another.
Use Triggers Instead of Constant Messaging
Behavioural triggers can help retailers communicate when something meaningful happens.
Possible triggers include:
- Product viewed repeatedly
- Basket abandoned
- Product back in stock
- Price changed
- Customer reaches loyalty threshold
- Expected replenishment date approaching
- Customer becomes inactive
A trigger can often produce a more relevant interaction than simply sending another scheduled campaign.
Dotdigital’s current automation guidance, for example, shows how product-view and cart data can be used to re-engage browsers automatically with relevant products and supporting content.
Dotdigital – Automated Product Recommendations – https://marketing.help.dotdigital.com/en/articles/12890304-convert-browsers-into-buyers-with-automated-product-recommendations
Operational Insight
Ask:
“What happened that makes this communication useful now?”
If there is no good answer, the message may not need to be sent.
Use Personalised Offers Carefully
Offers are another powerful personalisation tool.
But there is an obvious commercial risk: giving discounts to customers who would have purchased anyway.
Rather than presenting the same promotion to everyone, retailers may use information about:
- Customer intent
- Basket value
- Previous purchases
- Loyalty
- Price sensitivity
- Abandonment behaviour
…to determine when an incentive is appropriate.
RevLifter specialises in this area, using behavioural and intent signals to personalise offers and promotions.
RevLifter – https://www.revlifter.com/
Its current analysis highlights the margin problem associated with indiscriminate promotions: if shoppers learn that discounts are always available, incentives can become an expected part of the purchase rather than something that changes behaviour.
RevLifter – Customer Intent and Promotions – https://www.revlifter.com/blog/how-customer-intent-makes-promotions-work-harder-while-protecting-your-margins
Commercial Principle
The question is not:
“Did the personalised offer convert?”
It is:
“Would that customer have converted without the offer?”
That distinction matters enormously when measuring value.
Don’t Confuse Personalisation with Discounting
Personalisation does not need to involve financial incentives.
Often the most valuable intervention is simply making the experience more relevant.
Examples include:
- Showing preferred categories
- Improving search results
- Remembering previous browsing
- Highlighting available sizes
- Showing compatible products
- Providing useful delivery information
- Removing irrelevant messages
This can improve customer experience without sacrificing margin.
Retailers should therefore treat discount personalisation as one tool within a much broader strategy.
Avoid the Personalisation ‘Creepiness’ Problem
There is a point at which useful relevance becomes uncomfortable.
Customers may appreciate:
“Here’s something similar to what you looked at yesterday.”
They may respond differently to messaging that exposes unexpectedly detailed assumptions about their behaviour.
Personalisation teams should consider:
- How obvious the data use is
- Whether customers reasonably expect it
- Whether sensitive characteristics are being inferred
- How frequently targeting occurs
- Whether the customer has control
The ICO’s current data-minimisation guidance states that organisations should ensure personal data is adequate, relevant and limited to what is necessary for the stated purpose.
ICO – Data Minimisation – https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-protection-principles/a-guide-to-the-data-protection-principles/data-minimisation/
That provides a useful practical rule for personalisation: Don’t collect or use data simply because the technology makes it possible.
Use information because it creates a clearly defined customer or business benefit.
Get Consent and Data Governance Right
Personalisation teams need to work closely with privacy, legal and data-governance colleagues.
Depending on the technology and data involved, considerations may include:
- Cookies
- Tracking
- Consent
- Purpose limitation
- Data minimisation
- Retention
- Customer rights
- Third-party processors
The UK’s data-protection landscape has also changed following the Data (Use and Access) Act 2025/26, and the ICO is updating relevant guidance accordingly.
That makes it important for retailers to work from current regulatory guidance, rather than assuming historic implementations remain appropriate indefinitely.
ICO – Data Protection Principles – https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-protection-principles/
First-Party Data Makes Personalisation More Valuable
Retailers increasingly benefit from information customers generate directly through their relationship with the business.
Examples include:
- Purchases
- Account information
- Loyalty activity
- Product preferences
- Website behaviour
- Customer-service interactions
This first-party information can provide a strong foundation for personalisation because it reflects the retailer’s actual customer relationship.
But data quality matters.
Duplicate records, outdated preferences and fragmented identities can undermine personalisation.
Before implementing highly sophisticated journeys, retailers may need to improve:
- Customer identity resolution
- Product taxonomy
- Data synchronisation
- Consent records
Technology partners including Alchemy Network, Arsenalia, CTI Digital, Merx, Magebit and On Tap can support this broader digital architecture alongside personalisation itself.
Don’t Try to Personalise Everything at Once
Personalisation programmes can become unnecessarily complicated.
Imagine a retailer defining:
- 40 customer segments
- 20 behavioural triggers
- 12 channels
- 50 recommendation rules
The theoretical number of possible experiences quickly becomes enormous.
Instead, begin with high-value use cases.
A practical sequence might be:
1. Product recommendations
2. Basket abandonment
3. Browse abandonment
4. Post-purchase suppression
5. Cross-selling
6. Retention
Each use case can be tested before greater complexity is added.
Implementation Principle
Earn the right to become more sophisticated.
Get simple personalisation working reliably before creating dozens of complex journeys.
AI Can Help Scale Personalisation
AI and machine learning can help retailers analyse customer and product data at a scale difficult to manage manually.
Applications include:
- Recommendation engines
- Predictive segmentation
- Next-best action
- Offer optimisation
- Customer-intent modelling
- Content selection
But AI does not remove the need for commercial rules.
A model may identify a product as statistically relevant while failing to understand:
- Brand priorities
- Stock constraints
- Margin
- Customer sensitivity
- Current promotions
Retailers should therefore understand where algorithms operate autonomously and where business teams can apply controls.
Personalisation Needs Operational Ownership
One overlooked question is: Who actually owns personalisation after launch?
Responsibility may sit across:
- eCommerce
- CRM
- Marketing
- Merchandising
- Data
- Technology
- If everyone owns part of it but nobody owns the overall journey, personalisation can become fragmented.
Businesses should establish clear ownership for:
- Strategy
- Data
- Campaigns
- Product rules
- Testing
- Performance
Agencies and specialist partners can provide implementation and optimisation support, but internal ownership remains important.
Test Every Personalised Experience
Personalisation should be treated as a hypothesis.
For example:
Hypothesis: Showing complementary products after an item is added to basket will increase average order value.
That can then be tested.
A/B testing can compare:
Control: Existing experience
Variant: Personalised experience
Useful measures might include:
- Conversion
- Average order value
- Revenue per visitor
- Click-through
- Repeat purchase
Without a control group, retailers risk attributing sales to personalisation that would have happened anyway.
Measure Incremental Value
One of the biggest personalisation measurement mistakes is relying on attributed revenue.
Imagine:
Customers who saw personalised recommendations generated £500,000 in sales.
That sounds impressive.
But perhaps they would have generated £480,000 without personalisation.
The incremental value is closer to £20,000 than £500,000.
That is why testing matters.
Measurement Principle
Don’t ask:
“How much revenue interacted with personalisation?”
Ask:
“How much additional value did personalisation create?”
That is the figure that should ultimately inform ROI.
Look Beyond Conversion Rate
Personalisation can influence more than immediate purchase.
Potential metrics include:
- Product discovery
- Conversion
- Average order value
- Basket abandonment
- Repeat purchase
- Retention
- Customer lifetime value
- Unsubscribe rate
- Margin
Retailers may also monitor negative signals.
For example, a personalised email campaign that produces additional revenue but dramatically increases unsubscribes may have a longer-term cost.
Performance should therefore be considered across the wider customer relationship.
Personalisation and Fulfilment Need to Connect
Customer experience does not stop when the checkout closes.
Personalised promises around delivery and availability need to reflect operational reality.
This makes logistics information increasingly relevant.
A personalised message recommending an urgent purchase loses value if the product cannot actually be delivered when required.
Fulfilment providers such as Culina illustrate the operational layer beneath eCommerce customer experience: warehousing, inventory and distribution information can ultimately influence what retailers should promise customers.
Personalisation should therefore avoid operating in isolation from:
- Inventory
- Fulfilment
- Delivery
Greater relevance requires accurate operational information as well as customer data.
How to Build a Practical eCommerce Personalisation Strategy
A useful implementation model is:
1. Identify the customer problem
Where is greater relevance likely to improve the journey?
2. Define the desired outcome
Conversion, discovery, retention, margin or customer experience?
3. Identify the minimum necessary data
Which signals are genuinely required?
4. Design the personalised intervention
Recommendation, message, content or offer?
5. Connect the required systems
Can customer and product information update quickly enough?
6. Establish control
Which merchandising, privacy and commercial rules apply?
7. Test against a control
Did the personalised journey actually perform better?
8. Learn and expand
Scale successful use cases rather than adding complexity for its own sake.
This approach keeps personalisation connected to measurable problems.
Frequently Asked Questions
What is an eCommerce personalisation strategy?
It is a structured approach to using customer, behavioural and product information to make shopping experiences more relevant across websites, messaging and other customer channels.
What customer data can be used for personalisation?
Depending on the use case and appropriate data-protection basis, retailers may use information including browsing behaviour, purchase history, preferences, basket activity and customer status.
What is behavioural personalisation?
Behavioural personalisation adapts experiences according to actions customers take, such as viewing products, searching, abandoning a basket or making a purchase.
How can retailers personalise without offering discounts?
Product recommendations, relevant search results, remembered preferences, useful content, delivery information and post-purchase messaging can all create personalised experiences without discounting.
Does eCommerce personalisation require AI?
No. Many useful experiences can be created through rules and segmentation. AI can help automate and scale more sophisticated recommendation and prediction capabilities.
How should personalisation ROI be measured?
Where possible, retailers should use controlled experiments to measure the incremental impact on outcomes such as conversion, average order value, margin and retention.
Can too much personalisation damage customer experience?
Yes. Poorly targeted, repetitive or overly intrusive personalisation can reduce trust and make the experience less useful rather than more relevant.
Product Guide
Implementing eCommerce personalisation can involve marketing automation, digital experience, customer data, development, optimisation and fulfilment expertise. The following providers support different elements of that ecosystem, and can be met at the eCommerce Forum:
Featured Suppliers
Alchemy Network Ltd
Digital and technology partner supporting organisations with transformation, customer experience and platform integration, helping businesses connect the technology and data required to create more joined-up digital journeys.
Website: https://www.alchemy-network.co.uk/
Arsenalia
International digital-transformation group with specialist capabilities spanning customer experience, commerce, technology and data, supporting brands in creating integrated and personalised digital experiences.
Website: https://www.arsenalia.com/en
Brevo
Customer relationship and marketing platform combining email, SMS, automation, segmentation and customer-data capabilities, enabling businesses to create targeted communications based on customer characteristics and behaviour.
Website: https://www.brevo.com/
CTI Digital
UK digital agency specialising in digital experience, commerce, development and integration, helping organisations improve customer journeys and connect the platforms underpinning personalised experiences.
Website: https://www.ctidigital.com/
Culina
UK logistics and supply-chain provider supporting brands and retailers with warehousing, distribution and fulfilment. Accurate inventory and delivery information can provide an important operational foundation for personalised eCommerce promises and post-purchase experiences.
Website: https://www.culina.co.uk/
Dotdigital EMEA Ltd
Customer-experience and marketing automation provider offering behavioural targeting, dynamic product recommendations, segmentation and cross-channel communications across web, email, SMS and mobile.
Website: https://dotdigital.com/
ICS-digital
International digital marketing specialist supporting brands with multilingual content, search, digital strategy and audience engagement across global markets.
Website: https://www.ics-digital.com/
Magebit
eCommerce agency specialising in platform development, integrations and optimisation. Its work helps retailers build the technical environment required to connect commerce, customer and product data.
Website: https://magebit.com/
Merx
Digital commerce specialist supporting organisations with commerce technology, customer-experience design and optimisation around individual business requirements.
Website: https://www.hellomerx.com/
On Tap OÜ
eCommerce development specialist providing platform development, integrations, technical optimisation and ongoing support for online retailers.
Website: https://www.ontapgroup.com/
RevLifter
eCommerce personalisation specialist focused on intelligent offers and incentives. Its technology uses behavioural and intent signals to help retailers determine which promotions to offer, to whom and when, with an emphasis on improving conversion without unnecessary discounting.
Website: https://www.revlifter.com/
Sozo Design
Digital agency providing web design, eCommerce development and digital marketing services, supporting businesses in developing and improving online customer experiences.
Website: https://sozodesign.co.uk/
VINSINFO
Technology and digital-solutions provider supporting organisations across software development, commerce and digital transformation, including integration and development work underpinning customer-facing digital experiences.
Website: https://vinsinfo.com/
Turn Personalisation into Relevance
The real objective of eCommerce personalisation is not to create a different website for every customer.
It is to make each interaction more appropriate to what the customer is trying to do.
That often means beginning relatively simply: recognise meaningful behaviour, connect customer and product data, remove obviously irrelevant experiences and test whether each intervention genuinely improves the journey.
Over time, recommendation engines, automation and AI can add greater sophistication.
But sophistication should never become the objective in itself.
The best personalisation may be almost invisible to the customer.
They simply find the right product more quickly, receive the right information at the right moment and encounter fewer messages that have nothing to do with them.
The eCommerce Forum connects senior eCommerce, digital and payments professionals with carefully selected technology and service providers through a programme of pre-arranged one-to-one meetings.
Explore eCommerce personalisation solutions and specialist partners to discover how customer data, behavioural signals and technology can be turned into more relevant (and more commercially effective) digital experiences.
Related Reading
This article follows our earlier guide to eCommerce Personalisation Software: What Retailers Should Compare, covering customer data, recommendations, behavioural targeting, automation, privacy, integrations and supplier selection.
Sources
- Information Commissioner’s Office – Data Minimisation – https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-protection-principles/a-guide-to-the-data-protection-principles/data-minimisation/
- Information Commissioner’s Office – Data Protection Principles – https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-protection-principles/
- Dotdigital – Personalisation – https://dotdigital.com/personalization/
- Dotdigital – Complete Guide to eCommerce Personalisation – https://dotdigital.com/blog/the-complete-guide-to-ecommerce-personalization/
- Dotdigital – Recommendation SmartBlocks – https://personalization.help.dotdigital.com/en/articles/8523414-recommendation-smartblocks
- Dotdigital – Automated Product Recommendations – https://marketing.help.dotdigital.com/en/articles/12890304-convert-browsers-into-buyers-with-automated-product-recommendations
- RevLifter – https://www.revlifter.com/
- RevLifter – Customer Intent and Promotions – https://www.revlifter.com/blog/how-customer-intent-makes-promotions-work-harder-while-protecting-your-margins
- Brevo – Marketing Platform – https://www.brevo.com/products/marketing-platform/
Image credit: https://unsplash.com/photos/black-and-gray-laptop-computer-HTDVSbFsy3U
