01
CRO audit
Analytics, funnel drop-off, speed, accessibility and checkout reviewed against evidence, then a prioritised list of what to test or fix.
Elgrith London · CRO
Research it. Test it. Ship it. CRO audits, A/B tests set up within the ICO’s guidance, and every winning change built into your site by the same team, with a London client partner.
- New Delhi · HQ--:--
- New York--:--
- London--:--
- Shanghai--:--
FIG 02 · Quick answer
What does Elgrith do as a CRO agency in London?
Elgrith is a CRO agency serving London and UK brands through our London client-partnership centre, with engineering at our New Delhi HQ. Our conversion rate optimisation services cover CRO audits, A/B tests set up within the ICO’s consent guidance, Shopify and checkout optimisation, and building every winning change into your site.
- Projects delivered since 2015
- 1,000+
- Clients served worldwide
- 250+
- Tests set up within its consent guidance
- ICO
- Fixed written quotes
- GBP
Why it matters now
Most visitors leave. The rules on testing them changed.
Conversion work starts with evidence, not a list of best practices. These are the facts we plan around, checked on each source’s own page on 23 Oct 2026.
01
70.22%
Average documented cart abandonment
Baymard Institute’s average across 50 studies of online shopping carts. Most of those shoppers were close to buying, which is why checkout is where we look first.
Source: Baymard Institute, cart abandonment rate (last updated 22 Sep 2025)(opens in a new tab)02
A/B tests
Listed by the ICO as a likely statistical purpose
The ICO’s storage and access guidance lists A/B testing that compares groups of users as likely to fit the statistical-purposes exception, if conditions are met. Recordings of individual visitors still need consent.
Source: ICO, What are the exceptions? (storage and access technologies guidance)(opens in a new tab)03
30 Sep 2023
Google Optimize was switched off
Google says Optimize has been “no longer available as of September 30, 2023”. Many UK sites still carry its old snippet or have no testing set-up at all since.
Source: Google Optimize Help, sunset notice(opens in a new tab)
What you get
Conversion rate optimisation services for London brands
Research, testing and the build, from one team, so a winning idea reaches your live site instead of sitting in a slide deck.
02
A/B testing
Hypotheses, test design, sample size and test length agreed up front, QA on every variant, and results read against the threshold we set before launch.
03
Checkout and basket
Fewer steps, clearer delivery costs, VAT-inclusive prices, guest checkout and the payment options UK shoppers expect, measured from basket to order.
04
Landing pages
Campaign and product landing pages designed around one action, built fast and tested against your current page where traffic allows.
05
Shopify CRO
Shopify product, collection, cart and checkout improvements, built into your theme rather than added through a stack of apps that slow the store down.
06
Lead-gen and B2B forms
Enquiry, demo and quote forms with fewer fields, clearer errors and the follow-up wired to your CRM, measured as GA4 key events.
07
Speed and Core Web Vitals
LCP, INP and CLS improved in the templates that carry your revenue, because slow pages lose buyers before any test can help.
08
Accessibility
WCAG 2.2 AA issues in forms, buttons and checkout fixed as part of the work, so every visitor can complete the journey you are optimising.
Deliverables
10
deliverables, in writing
Included in every build, written into your fixed quote before work starts.
Test or fix?
Is there enough traffic to A/B test?
Not every site should run A/B tests. Before we propose one, we calculate the sample size each variant needs to detect the smallest change worth having, then check whether your traffic gets there inside a sensible test window.
| Your situation | What we recommend | Why |
|---|---|---|
| Enough conversions to reach the sample size within the agreed test window | A/B test the change | You get a measured answer before rolling it out to everyone |
| Too few conversions to reach that sample size in time | Research and fix | Evidence-backed changes shipped, then the trend watched, instead of a test that can’t conclude |
| Tracking broken or key events missing | Fix measurement first | Without reliable GA4 key events, neither a test nor a fix can be judged |
| The problem is price, product or delivery | Say so plainly | CRO can’t fix an offer buyers don’t want; we tell you before you spend on testing |
We do not use a blanket traffic threshold: the answer depends on your conversion rate, the size of change you care about and how long you can run a test. It is worked out in the audit and written into your scope.
Process · 6 steps
How we work with London teams.
A London client partner, reviews in your working day, and a written test plan before anything goes live. You own the data, the code and the IP.
- Step 01
Free call
Your goals, traffic, platform, tracking and what you have already tried, discussed by video in your working day.
- Step 02
CRO audit
Analytics, funnel, speed, accessibility, checkout and consent set-up reviewed, with findings ranked by evidence and effort.
- Step 03
Fixed written quote
A written scope with the test or fix backlog, how results are judged and a fixed quote in GBP.
- Step 04
Test or fix in sprints
Variants designed, built and QA’d, or evidence-backed fixes shipped, with a weekly review of what is live.
- Step 05
Build the winner
Winning variants rebuilt properly in your theme or codebase, not left running inside a testing tool.
- Step 06
Report and next test
A plain-English write-up of every result, including the losers, and the next items in the backlog.
What you get
Included in every CRO engagement
No prices on this page, on purpose. Every project gets a written scope and a fixed quote in GBP after a free call, so you know the full cost before work starts.
10items
- Written scope and fixed quote in GBP
- CRO audit with findings ranked by evidence
- GA4 key events and tracking plan checked
- Consent set-up reviewed against the ICO’s guidance
- Sample size and test length agreed before launch
- QA of every variant on desktop and mobile
- Winning changes built into your site
- Core Web Vitals and WCAG 2.2 AA checks on changed templates
- Results write-up, including tests that lost
- Analytics, testing and code accounts in your company’s name
A/B testing and UK cookie rules
Test within PECR and the ICO’s guidance.
Since the Data (Use and Access) Act 2025, the ICO’s guidance on storage and access technologies lists A/B testing that compares groups of users as likely to fit the “statistical purposes” exception, so it may run without consent when the conditions are met: the data is aggregate, individuals are not tracked, people get clear information and a free, simple way to object, and any third-party tool only helps you improve your service. Logs or recordings of individual visitors still need consent unless they are used for security. We set tests up on that basis, checked on the ICO’s page on 23 Oct 2026. This is not legal advice; your own adviser signs off your consent set-up.
- Aggregate A/B tests configured to the exception’s conditions
- Session recordings and heatmaps of individuals only after consent
- A clear notice and a simple, free way to object
- Third-party tools set up to process data for you only
- Consent banner checked before any test goes live
Working with a London team
London partner. UK working day.
A client partner at our London client-partnership centre runs your account on UK time. Your audit, tests and builds are engineered at our New Delhi headquarters, which is 4.5 hours ahead of London in summer and 5.5 hours ahead in winter, so our working day overlaps your morning.
A London client partner
One person owns your account from the London centre: scope, priorities, reviews and sign-off. You never chase a ticket queue.
Calls in your working day
Kick-offs, reviews and demos are booked in your working day, by video. Written updates follow every call.
New Delhi engineering
Analysis, design, development and QA happen at our New Delhi headquarters, the team behind 1,000+ projects since 2015.
In person if agreed
Sessions run remotely by default. If your project needs an in-person workshop in London, we agree it in the scope first.
New Delhi · HQLondon
Live: today’s working days in London and New Delhi, adjusted for British Summer Time automatically.
Your team
London partnership. New Delhi engineering.
Client partnership
London
UK & Europe · Client partnerships
--:--GMT/BST (UTC+0/+1)
Headquarters
New Delhi
Headquarters · Design & engineering
--:--IST (UTC+5:30)
Elgrith in London
A CRO agency for London and the UK
From ecommerce brands to B2B firms, UK teams use Elgrith to find what stops visitors converting and to build the fix. These are the areas we serve; we do not claim a London address.
01
Ecommerce and DTC brands
Product pages, basket and checkout on Shopify, WooCommerce, Magento or custom builds, with GBP and VAT-inclusive prices shown the way UK shoppers expect.
02
B2B and lead generation
Demo, quote and enquiry journeys for professional services, software and trade suppliers, where one extra form field can cost a lead.
03
SaaS and subscriptions
Pricing pages, sign-up and onboarding flows, measured from first visit to activated account as GA4 key events.
Serving businesses across London and beyond
- City of London
- Shoreditch
- Canary Wharf
- Soho
- Westminster
- Islington
- Greater London
- Across the UK, remotely
What we measure
The numbers we read before and after every change
Lift claims without a method mean little. We agree what is measured, how and for how long before a test or fix goes live.
07items
- GA4 key events (Google’s name for conversions since March 2024) for each step that matters
- Funnel drop-off from landing page to order or enquiry
- Basket and checkout abandonment, by device
- Core Web Vitals (LCP, INP and CLS) on the templates being changed
- WCAG 2.2 AA issues in forms, buttons and checkout
- Form errors and field-level drop-off
- A significance threshold and sample size set before each test starts
Questions to ask any CRO agency, including us
How to choose a CRO agency in London
Comparing CRO agencies in London? These questions separate an agency with a method from one with a slide of uplift numbers.
07items
- Will they show you their method: sample size, test length and significance threshold?
- Who builds the winning version, and does it leave the testing tool?
- Where does the team actually sit, and is that the address on their website?
- How do they handle cookie consent and session recordings under PECR?
- Will the analytics and testing accounts be in your company’s name?
- Do they report the tests that lost, not just the winners?
- Do they promise a lift before seeing your data? Treat that as a warning
Tech & tools
The CRO toolkit we work with
Measurement and build tools your team can keep using after we finish. We recommend testing tools; you hold the accounts.
- GA4
- Google Tag Manager
- Consent Mode
- BigQuery
- Looker Studio
- PageSpeed Insights
- Server-side testing
- Feature flags
- Figma
- Shopify
- Next.js
- WordPress
FAQ
Frequently asked questions
What London businesses ask us about conversion rate optimisation.
It depends on how many templates and journeys are in scope, whether you need a full CRO audit first, which testing tool you use, your traffic and how much build work each change needs. We do not publish prices: after a free call you get a written scope and a fixed quote in GBP.
Let’s find what stops your visitors converting.
Book a free call. Tell us your platform, traffic and the journey that matters most, and we will send a written scope with an audit or test plan and a fixed quote in GBP.
Free 30-minute call · London
- Fixed written quote, no surprises
- NDA before discovery
- You own the code and IP
Book a free call
OpenPick a day. You'll choose a time in your own time zone on the next step.