Skip to content
For Agencies Talk to Us

AI recommendations and personalisation
your customers will welcome.

An AI recommendation engine for on-site content, products and email segments. It is shaped by what your customers really do. It uses only data they agreed to share on your own site and systems. From day one we measure it against a holdout group (customers who never see it). So any lift is your own number, not our brochure's.

Get a ProposalGet a Free AI Automation Review

Reply within four business hours (typical) · Measured against a control group

Brands our teams have delivered for

  • Hello Peter
  • Gruber Logistics
  • Delhivery
  • Thomson Reuters
  • Qatar Airways
  • Grundfos
  • Save
  • BERD
  • Yale University
  • Kuwait Police
  • Dubai Police
  • Panasonic
  • Infosys
  • Kia
  • Hitachi
  • Orange Business Services
What we do

PixelCrayons builds AI personalisation on the tools you already run: content and product recommendations, smarter search results and email segments. It uses only data customers agreed to share with you, within written privacy limits. Every result is measured against a holdout group that never sees personalisation. We work direct for brands and white-label for agencies.

  1. Built on data you're allowed to useWhat customers did on your own site and channels, with their consent checked every time.
  2. On-site content & product personalisationContent blocks, product recommendations and smarter search results built into the store or CMS you run.
  3. Email and SMS segmentsGroups of customers your email and SMS flows can act on, planned with our email team.
Our work

See what we've built
and the results it brought.

Real projects for real clients, with results we can show.

All case studies →
2004
Established
500+
Agencies served
4,500+
Projects delivered
340%
Revenue growth · 7 months
eCommerce
+127%
Organic traffic · 5 months
SaaS
85%
Faster delivery · backlog cleared
via agency partner
Clutch rating 4.9 / 5, checked 15 September 2026Read our GoodFirms reviews
Google Partner
Shopify Partner
ISO/IEC 27001 certification seal (QRO), held by the Vinove groupISO 9001:2015 certification seal (QRO), held by the Vinove groupCMMI Level 3, held by the Vinove group
Technologies we work with
Commerce and web
  • Shopify
  • WooCommerce
  • Magento
  • WordPress
  • Webflow
  • Next.js
Marketing
  • Google Ads
  • Meta Ads
  • GA4
  • Klaviyo
Cloud
  • AWS
  • Microsoft Azure
  • Google Cloud
AI models
  • OpenAI
  • Anthropic
  • Google Gemini
Build and deploy
  • Docker
  • Kubernetes
  • Terraform
  • GitHub Actions
CRM and automation
  • HubSpot
  • Salesforce
  • Zapier
  • n8n

Product names and marks are trademarks of their owners. Naming a technology means our teams build with it. It does not mean a partnership, certification or endorsement.

  • Consented, first-party data only
  • A control group in every project
  • No invented lift figures: your baseline decides
  • Reply within four business hours (typical)
  • Sensitive categories excluded from targeting
  • Your stack: no forced platform
In every build

What a personalisation build
actually includes.

01

Built on data you're allowed to use

Personalisation uses what people browsed, bought and searched on your own site, with their consent. We never use bought third-party profiles or data your privacy policy doesn't cover. Consent is checked each time the system makes a choice. A visitor who declined tracking simply sees the standard page, with no tricks to nudge them back.

Every build
02

On-site content & product personalisation

Content blocks, personalised product recommendations and reordered search results on the site you already run. Returning customers see reorder shortcuts and categories they like. First-time visitors see your strongest general offer. It is built into your existing store or CMS, so you don't move platforms. Every rule can be checked, so your team can see exactly why a block appeared.

The surface
03

Email and SMS segments

The same signals group customers for your email and SMS flows: when they're due to reorder, which categories they like and who may stop buying. So your messages stop being one blast to everyone. We plan this with our email team, including Klaviyo builds. Your site and your emails then tell one story instead of two.

Included
04

Limits that keep it from feeling intrusive

Being relevant earns trust, but feeling watched loses it. Every build comes with its limits written down. Sensitive categories are never targeted. Nothing is based on traits people didn't knowingly share. There are caps on how often people see a message, and the wording never reveals how much the system knows. Our test is simple: if a customer saw exactly why they were shown something, would they mind? If yes, we don't build it.

Every build
05

Measured against a control group

A slice of your audience never sees personalisation, on purpose. Every result is measured against that control group, on your own numbers: conversion, order value and repeat rate. You'll notice this page quotes no average uplift. That's our policy, not an oversight. If the control group shows a tactic isn't paying, the report says so and we stop it.

Every report
The pilot

From a consent check
to measured results.

01
Start

Brief and access

We sign an NDA. We map the customer data you have and what your consent form allows.

02
Once the brief is complete

Audit and itemised proposal

We confirm the date your proposal arrives. It sets out your starting numbers and a priced scope for one part of your site: its segments, limits and control group size.

03
Your decision

You sign off

You confirm the scope, the payment and the access we need. Nothing starts before this.

04
Typically 2 to 3 business days later

The work starts

We build personalisation on one real part of your site, with consent checked and limits in place. The control group is kept apart, so measurement starts clean.

05
On the date in your proposal

First deliverable

The pilot goes live and results are compared with the control group. What pays is extended. What doesn't is stopped, in writing.

Inside Prism

Your engagement, week to week,
in one workspace.

Prism is the client workspace where your requests, approvals, tasks and weekly reviews live.

Some engagements include Prism, our client workspace, and your proposal says whether yours does. Where it does, your AI build runs in a workspace you log into: requests, approvals, the task list and the weekly review, with each decision and its expected result written down.

See how Prism runs an engagement
  • 01

    Requests and approvals

    One queue, one owner, one due date.

  • 02

    Weekly review

    Each decision recorded with its expected result.

  • 03

    Actions checked against outcome

    What we did and what happened, side by side.

What your data can support

You can only personalise
on what you really know.

Every personalisation tactic needs some data about the visitor. Visitors arrive in one of four states. Line the tactics up against them and most need a visitor you haven't met yet. So the fix is in how the work is planned, not in the technology.

What you can personalise for each type of visitor
Visitor stateEntry message & offer framingProduct & content rankingPricing & plan emphasisRepeat-purchase emails & retention
First visit, anonymousReferrer, campaign, search term, geography, device, time of dayWhere most first-time visitors areSupportedMatch the page to the campaign or search that brought them. It's cheap, and it's the most valuable thing most sites aren't doing.PartlyLocation and device only. “People like you” needs to know who you are first.No signalNo signal
Returning, anonymousThe above, plus what they viewed this session and lastSupportedPick up where they left off rather than restarting the pitch.SupportedWhat they do is now real data: viewed, compared, abandoned.PartlyGuessed from what they looked at. That's worth testing, but it isn't a fact.No signal
Known: logged in or identifiedThe above, plus account, plan, order history, entitlementsSupportedSupportedSupportedTheir current plan and usage mean you know which plan to stress, instead of guessing.PartlyPossible, but repeat-purchase messages belong in email and your CRM more than on the page.
Active customerThe above, plus usage, support state, renewal dateSmallest audience, most dataSupportedSupportedSupportedSupportedMessages about renewal, upgrades and the risk of leaving, planned with your email programme instead of repeating it.
  • Supported by real data
  • Partly: rough data only
  • No data: do not fake it

Start with the top row. Most first-time visitors are in the state with the least data, and nearly every proposal skips it.

Illustrative matrix: signal availability by visitor state, not measured performance

How we report it, in Prism

The holdout report,
before anyone quotes a lift.

What you receive in month one

Personalisation is reported against a control group that never sees it. In the first month you receive the consent audit naming which signals your privacy policy allows and which it does not, the holdout design with its size and assignment written down, and the baseline for the metrics you chose. Each rule is then read against the control, by segment, and a rule that does not beat it is retired.

  • 01Consent audit: the signals you may use, the ones you may not, and the boundary written down
  • 02Holdout design: the control group, how it is assigned, and what it never sees
  • 03Baseline: the metrics you chose, measured before any rule goes live
  • 04Lift by segment: each rule against the control, with the confidence stated
  • 05Retirement log: the rules that did not beat the control, and when they came out

Related workAl Majed for Oud: a fragrance house's whole catalogue, online in two languages.Commerce · Saudi Arabia · A different discipline, so it is linked here rather than presented as proof of this one.

Al Majed for Oud's Saudi store in Arabic on desktop: a campaign banner for the Serious perfume, with the brand's Arabic and English wordmark in the header
Al Majed for Oud’s site, Internet Archive copy, 1 April 2026
Our other AI and email services

Other services that work with personalisation

The data that shapes a page also shapes your customer emails. Our email team runs that side, Klaviyo included. Start where the problem is, or start at the AI page.

Email MarketingKlaviyoAll AI solutions
Who is accountable

The people responsible
for your work.

Leadership and how we’re built →
Parvessh Agarwal

Parvessh Agarwal

Founder & CEO

An entrepreneur at heart. He links business analysis, strategy, clear goals and ROI to practical technology.

LinkedIn →
Saundaraya Gupta

Saundaraya Gupta

Technical Manager

Brings technical work, client updates and project management together, so work moves smoothly from brief to launch.

LinkedIn →

Every engagement

A named coordinator and escalation contact

Know who coordinates your engagement, who owns the work and how to escalate an issue.

All engagements. Roles may be combined and vary with your engagement.

Full terms →
Before you sign

Our commitments

Each one says where it applies and links to its full terms.

Response and start times

We typically respond within four business hours. Eligible engagements can typically start within two to three business days after scope, payment and required access are confirmed.

Business hours are Monday to Friday, 9:00 am to 6:00 pm IST, excluding published company holidays. Typical expectations, not guaranteed service levels.

Full terms

Clear IP & Ownership Terms

Your materials remain yours. Bespoke deliverable rights, licences and handover are agreed before work starts.

All engagements.

Full terms

Clear Delivery & Escalation Ownership

Know who coordinates your engagement, who owns the work and how to escalate an issue.

All engagements. Roles may be combined and vary with your engagement.

Full terms

How we handle claims, credentials and your data: group credentials, dated ratings, how case figures are signed off, client permissions and data handling.

Rule by rule

Build, extend or stop?
How we decide on each rule.

Every personalisation rule is judged twice. Before it is built, we check it against your consent and your data. After launch, we compare it with a control group that never sees it. These are the choices we make at each point.

What we findWhat we doWhy
The signal a tactic needs is not covered by the consent you holdWe do not build it, and say so plainlyPersonalisation only uses data your consent form allows. The audit also shows what fixing your consent form would make possible.
Traffic or purchase history is smallStart with a few customer groupsA smaller plan helps more than pretending we can predict each person. The pilot scope says so.
A customer who saw why they were shown something would mindIt does not shipHelpful personalisation uses what people knowingly shared. Showing that you guessed something goes too far.
The personalised group beats the holdout on the metrics you choseExtend the rule to new surfaces and segmentsThe lift is your own number, compared with your own control group instead of a vendor average.
The rule does not beat the control group, or only moves revenue between productsStop it, in writingThe control group shows tactics that move revenue around instead of adding it. We record them as stopped instead of defending them.

This is why no page of ours quotes an average uplift: the control group decides.

Questions

Frequently
asked.

It can be, when it shows you guessed something a customer never told you. Helpful personalisation uses what they knowingly shared to save them effort. We never target sensitive categories or mention guessed traits in the wording, and we check consent every time. If a customer would mind seeing why they were shown something, it doesn't go live.

What customers did on your own site and channels: pages browsed, products bought, searches run and emails opened. It must be tied to consent you really collected. We don't buy third-party profiles. If your traffic or history is small, we start with a few customer groups. The discovery audit shows what your data supports before anything is priced.

Against a holdout: a random slice of your audience that never sees personalisation. Every result compares the personalised group with that control group on your own numbers: conversion, order value and repeat purchase. That's why we quote no average uplift. It's also why tactics that only move revenue around are stopped in the report.

Personalisation only uses data your consent form allows, and consent is checked every time. A visitor who declines tracking sees the standard page. If they withdraw consent later, personalisation stops. Data stays in your systems and how it is used is written down for your data protection officer. We'll tell you if a request doesn't fit the consent you hold.

They work together. An A/B test finds the version that's better for everyone. Personalisation shows different versions to different groups, and each rule is checked against the control group. Test results show which groups are worth personalising for. If you already run a conversion programme, we plug into it instead of repeating it.

It depends on where it runs, the data work and the limits your sector needs, so we don't quote from a rate card. Every project starts with a pilot on one part of your site, priced as a fixed build. Adding more pages or email flows is priced step by step, once the control group numbers justify it.

Both. On a store it recommends products, reorders search results and shapes content for returning customers and first-time visitors. The same data, used with consent, groups customers for your email and SMS flows. It is built into the store or CMS you already run, so you don't move platforms.

The range

What it costs,
at the direct rate.

First comes an audit of your customer data and consent, showing what it can honestly support. Then a build in one place, measured against a group that sees no changes, using your own numbers.

  • One-offData and consent audit$1,250–$3,750one-time
  • ProjectPersonalisation build$6,250–$31,250fixed quote, one-time

Indicative: every figure is refined into an itemised proposal.INR figures shown excl. GST.

PricingSee indicative pricing for this serviceDirect and wholesale, in ₹ and $ →

Personalise one surface,
and measure it honestly.

Scope a pilot: one surface, consented data, a preserved holdout, with the guardrails written down and results judged against your own baseline, not a brochure.

Get a ProposalGet a Free AI Automation Review

Reply within four business hours (typical) · NDA standard · Clear IP & Ownership Terms

Rankings on this topic

Scored on public evidence by the PixelCrayons Index. PixelCrayons is scored like every firm: ranked where it has a case study, listed where it has not.

Last updated