B2B CRO Agency for SaaS
As a SaaS CRO agency, we run research-led conversion rate optimisation every week, as part of Run in the Surge GTM Engine. Most SaaS conversion problems are not button colours. They are a demo form that asks fourteen questions, a pricing page that will not say a number, and a trial that takes four days to reach first value. We find where qualified traffic actually drops out, then rebuild that part of the journey around deal value rather than form fills.Load more
Conversion optimisationThe gaps quietly costing you pipeline.
01Traffic is growing and demo requests are flat, so the funnel is losing people somewhere nobody has located
- What it costs
- Spend keeps buying visits that leak out of the funnel, and nobody can say which stage to fix first.
- What we build
- A full-funnel drop-off map by channel and stage, so the first change goes where the leak actually is.
02Sales says the leads are unqualified, marketing says there are plenty of them, and both are looking at different numbers
- What it costs
- Teams argue about volume while qualified pipeline goes unmeasured, so no decision about the funnel can be trusted.
- What we build
- One measure both teams share, optimised for deal value rather than form fills, reported from your own CRM.
03The pricing page will not state a price, so buyers who cannot get one leave and are never counted
- What it costs
- Ready buyers quietly drop out at the point of decision, and the loss never shows in any report.
- What we build
- Research into why buyers leave the pricing page, then reasoned changes measured against qualified pipeline.
04Trial sign-ups are healthy but activation is not, and nobody has measured time to first value
- What it costs
- Acquisition spend creates trials that never convert, and the gap between sign-up and value stays invisible.
- What we build
- Activation measured as its own stage, with session research showing where new users stall before first value.
05Testing happens, but on low-traffic pages where no result could ever reach significance
- What it costs
- Months go on tests that cannot conclude, and the team ships changes based on noise.
- What we build
- Tests only where traffic lets them conclude; on smaller pages, reasoned changes with the reasoning written down.
06The demo form asks for information sales does not use, and every field costs completions
- What it costs
- Every unnecessary field costs completed demo requests from buyers who were ready to talk.
- What we build
- A form audit against what sales actually uses, removing the rest or moving it after the first conversation.

How we build, run and measure it.

A full-funnel drop-off map by channel and stage. Many conversion rate optimisation agencies start testing before they know where the loss is.
A change that raises form fills and lowers qualified pipeline is a loss. We measure the thing sales is judged on.
Below a traffic threshold an A/B test cannot reach significance. On those pages we make reasoned changes and say so, rather than performing statistics.
Session recordings, exit surveys and sales-call listening. The reason people leave is usually knowable and usually not what the team assumed.
Why SaaS conversion is a different problem
Ecommerce CRO optimises a single session towards a purchase. A SaaS buying cycle runs for months across many people and several systems, and that changes almost everything about how SaaS conversion rate optimisation services should be run.
- The conversion event you can see, a form fill, is months away from the revenue event that matters
- Multiple people at the same account convert separately, which corrupts a naive per-visitor calculation
- The highest-value change is often removing a step rather than optimising it
- Self-serve and sales-led motions on the same site pull the page in opposite directions
- Activation and time to first value belong to conversion, but usually sit with product and go unmeasured
- Test velocity is capped by traffic, so most SaaS sites can run far fewer valid tests than their agency proposes

Where the work usually is
Form and qualification design
Every field costs completions. We audit which answers sales actually uses, and remove the rest or move them after the booking.
Pricing page honesty
A pricing page that refuses to indicate a number loses the buyers who cannot get budget without one, and they leave without a trace.
Trial activation and time to first value
Measure how long it takes a new account to reach the moment the product becomes useful, then remove the steps before it.
Journey design for self-serve and sales-led
Two motions on one site need two paths. Forcing both through the same page loses one of them.
Page speed and mobile experience
Unglamorous, measurable, and still the largest fixable loss on a meaningful share of SaaS sites.
Qualitative research
Exit surveys, recordings and listening to sales calls. This is where hypotheses worth testing come from.
How a CRO engagement runs
Instrument and baseline
Establish what can currently be measured, fix the gaps, and record the drop-off at every stage before changing anything.
Locate the largest loss
Quantify each leak in pipeline terms. The biggest percentage drop is often not the biggest revenue loss.
Research why
Recordings, surveys and sales-call listening on the specific step that is losing people, rather than site-wide.
Fix the obvious first
Removing an unnecessary form field does not need an A/B test. We separate the changes that need proving from the ones that need doing.
Test what genuinely needs testing
Sized against your traffic, with the required sample declared before the test starts and the result honoured either way.
Report in pipeline, then repeat
Every change reported in deal value, and the drop-off map refreshed so the next largest loss is the next piece of work.
Measured in pipeline and revenue.
Built in your accounts, run with your team, reported against the numbers below.
How we measure it
- Visitor-to-lead conversion
- MQL-to-SQL conversion
Targets are agreed from your first month of data.
Conversion & Lifecycle Journey Map
Find and fix where pipeline drops out across the website, forms, demo/trial, nurture, onboarding and renewal, and run lifecycle programmes that convert, retain and expand.
In B2B SaaS small conversion gains compound across the funnel; retention and expansion drive NRR, which drives valuation.
- When we use it
- Run week 8 once traffic and lifecycle data are reliable.
- Typical timeframe
- 2 weeks to map; 4–6 week test cycles; quarterly lifecycle review
- What it needs
- Funnel data (F06), GA4 and session recordings, messaging (F03), product usage (PLG), CS health data.
- Channels
- Website, Email, in-app
- Tools, in your accounts
- GA4, Microsoft Clarity / Hotjar, VWO / Optimizely, Chili Piper, HubSpot / Customer.io, Vitally / Planhat
What we track
- Visitor → lead conversion
- GA4 · weekly
- MQL → SQL conversion
- CRM · weekly
- CRO tests shipped
- Test log · monthly
- Nurture → opportunity rate
- CRM · monthly
- Logo retention
- CRM / CS tool · quarterly
- Expansion pipeline £
- CRM · monthly
Targets are agreed from your first month of data.
Exactly what the team does.
Phase by phase, with the week each step lands. Week 1 is the start of Build.
01Week 8Map
- Journey map. Map each journey (demo, trial, pricing, content) with stage conversion rates. → Journey map
- Drop-off sizing. Size each drop-off in £ pipeline using the growth model. → Prioritised drop-offs
02Week 8–9Research
- Qualitative research. Session recordings, form analytics, 5 user interviews, sales feedback. → Insights
03Week 9 – FortnightlyTest
- Test backlog. Hypothesis, metric, impact score; prioritise with F16. → Backlog
- Run tests. Ship 2+ tests/month; bigger changes given B2B traffic; run to significance or 4 weeks. → Test results
- Demo/booking flow. Instant booking (Chili Piper), fewer fields, routing to right AE. → Booking flow
04Week 10–11Lifecycle
- Nurture programmes. By persona and stage; recycle and closed-lost re-engagement. → Nurture flows
05Week 12Retention
- Health and churn signals. Usage, support, NPS, champion change → risk alerts. → Health score
- Expansion plays. Usage thresholds, new teams, upsell triggers → CS/AE tasks. → Expansion plays
What we send you
- Journey map with £ drop-offs – where to focus. Week 8
- Test backlog and results – transparency. Fortnightly
- Nurture flow designs – approval. Week 10
- Retention and expansion playbook – cS adoption. Week 12
How we know it's done.
Definition of done
- Journeys mapped with conversion rates
- Drop-offs sized in £
- Backlog prioritised
- 2+ tests/month shipping
- Nurture and closed-lost flows live
- Health and expansion signals live
Risks we plan around
- Low traffic for tests
- Test bold changes; use qualitative evidence
- Nurture spam
- Frequency caps; stage-relevant content
- CS not engaged
- Co-design health score with CS
Why B2B SaaS teams choose us.
We locate before we test
Most CRO programmes start testing before establishing where the loss actually is.
Deal value, not form fills
The metric is qualified pipeline, agreed with sales before the work starts.
Honest about statistics
We will tell you when your traffic cannot support a valid test, rather than running one anyway.
Research-led hypotheses
Tests come from observed buyer behaviour, not from a list of best practices.
Joined to attribution
CRO reporting that reconciles with the CRM, so results survive a finance review.
We will say when it is not CRO
A qualified-traffic problem does not get fixed on the landing page, and we would rather redirect the budget.
Questions teams ask.
What is a good SaaS conversion rate?
There is no single good number. It depends on which stage you measure, where the traffic comes from and your ACV: visitor-to-signup, trial-to-paid and demo-to-opportunity are different rates, and published benchmarks often mix them. The useful comparison is your own funnel by channel and stage over time, measured against qualified pipeline rather than form fills. We baseline it in the first weeks and agree targets from your own data.
How do you improve a SaaS conversion rate?
Find the largest loss first, then fix it. Map drop-off by channel and stage, research why people leave that step through recordings, exit surveys and sales calls, remove obvious friction such as unused form fields, and test only what genuinely needs proving. Measure every change in qualified pipeline, so a lift in form fills that lowers deal quality shows up as the loss it is.
What firms specialise in CRO for SaaS companies?
Look for a B2B CRO agency that measures conversion in pipeline rather than form fills, works in your own analytics and CRM, names the senior person doing the work, and tells you when your traffic cannot support a valid test. Many CRO firms bring ecommerce habits to a SaaS funnel and miss at least one of those. Surge45 runs CRO that way as part of a GTM engine, but judge any shortlist, us included, on the same tests.
How do you A/B test a SaaS pricing page?
First check the page gets enough qualified traffic for a test to conclude, because many do not. If it does, test one change at a time, such as plan structure, a starting price or the call to action, declare the sample size before you start, and judge the result on trials, demos and pipeline rather than clicks. If traffic is too thin, make a reasoned change and compare before and after honestly.
How much traffic do we need for this to work?
For a valid A/B programme, considerably more than most SaaS sites have on their highest-intent pages. Below that threshold the work is research-led: locate the loss, understand why, make reasoned changes and measure the before and after honestly. We will tell you which category you are in at the start.
How long before we see results?
Instrumentation and the drop-off map take two to three weeks. Removing obvious friction can move numbers within a month. A test programme moves at the pace of your traffic, so anyone quoting a fixed timeline has not looked at it.
Do you guarantee a conversion uplift?
No, and be wary of conversion optimisation services for SaaS companies that do. Some tests lose, which is information rather than failure, and a programme that never reports a losing test is not reporting honestly. What we commit to is finding where the loss is, measuring it in pipeline terms, and reporting what each change actually did.
Should we put pricing on the site?
Usually at least an indication, yes. Buyers who need a number to start an internal budget conversation leave without contacting you, so the cost of hiding it is invisible in your analytics and real in your pipeline. Even for bespoke enterprise pricing, a starting band usually helps more than it hurts.
More ways we help.
Paid media
LinkedIn first, with Google Ads and programmatic as add-ons, run against pipeline rather than clicks.
Learn moreLead generation
Multi-channel lead generation that fills the pipeline with qualified buyers.
Learn moreLifecycle & nurture
Onboarding, nurture and lifecycle email run with your team.
Learn moreMeasurement & attribution
One pipeline number and attribution your CFO can reproduce.
Learn more