Customer journey analysis finds the one stage where you lose the most people, and tells you how long the rest take to get through. It turns a journey you have already defined into two numbers per stage: how many move on, and how many days it takes them. This article answers how to run that analysis with pipeline data on the levelpartner.ai platform, how to read the results without fooling yourself, and what to change first.
The two numbers that matter
Stage conversion is the share of people who move from one stage to the next. Time in stage is the number of days they spend there before moving or dropping out. Everything else in journey analysis is a refinement of those two.
Low conversion at a stage means people are leaving. Long time in stage means people are stuck. They call for different fixes. Leaving usually points at a mismatch: the wrong people, the wrong offer, or a poor experience at that step. Stuck usually points at a process problem: nobody follows up, documents are slow, a decision-maker is missing.
Why it matters for a business with flat revenue
Most firms with stalled turnover try to fix the top of the journey by buying more leads. Journey analysis usually shows the biggest loss is further down, at proposal, at document collection or at the second meeting. Lifting conversion at one mid-journey stage raises revenue from every lead you already have, without increasing spend.
It also prevents the opposite error: blaming marketing for lead quality when the leads are fine and the follow-up is slow. The numbers settle that argument quickly.
Before you start
- Pipeline stages that each mark a real event. If they do not, see customer journey stages first.
- At least three months of opportunities moving through the pipeline.
- Lost opportunities closed with a reason rather than left open.
- Stage changes made at the time they happen, not in a batch at month end.
How to set it up
- Close or archive stale opportunities older than two sales cycles.
- Go to Dashboard, click Edit Dashboard and add the Funnel widget.
- Set the date range to the last full quarter.
- Note the count of opportunities that entered each stage.
- Read the next-step conversion the Funnel shows for each stage.
- Count the deals your Stale Opportunities workflow has tagged in each stage.
- Mark the stage with the lowest conversion and the one with most stale deals.
- Pull ten lost deals from the weakest stage and read their timelines.
- Write one change to test at that stage over the next quarter.
Take 300 valuation enquiries in a quarter: suppose 180 book a valuation, 150 are held, 60 receive a proposal and 42 instruct. Held to proposal, at 60 of 150, would be the weakest step, and say the average time from valuation to proposal is nine days. If reading ten lost deals showed most vendors instructed a competitor who sent terms within 48 hours, the change to make would be a target of sending proposals the next working day. The numbers are illustrative.
Reading the results honestly
Look at stages with enough volume. A stage with eight opportunities will swing wildly from quarter to quarter and teaches you nothing. Group small segments together or use a longer date range.
Compare like with like. Leads from referrals and leads from paid social behave differently. If you can, run the analysis separately by source, using the source data from marketing attribution. A drop at proposal that only affects one source is a targeting issue, not a sales one.
The mistake most people make
Trying to fix every stage at once. Change three things in a quarter and you will not know which one worked. Pick the single weakest stage, make one change, measure it for a full sales cycle, then move on.
Time it takes
The first analysis takes two to three hours, mostly cleaning old opportunities. After that, a quarterly run takes under an hour. Reading lost deals is the valuable part and should not be skipped. If you want the stage numbers always visible, add them to a sales and marketing dashboard.
When not to use it
If your pipeline is not kept current, the analysis will describe your admin habits rather than your clients. Fix pipeline discipline first. And if you have only a few dozen opportunities a quarter, stage percentages will mislead; read individual deals instead.
Troubleshooting
Conversion appears higher than reality. Opportunities are being added only once a deal looks likely. Create opportunities at enquiry, not at proposal.
No deals ever show as stale at some steps. Staff move deals through several stages at once at the end. Ask the team to move deals on the day an event happens, or automate stage moves on bookings and form submissions.
Lost reasons all say "other". The reason list does not fit. Rewrite it with five specific options drawn from recent lost deals.
Results vary hugely by month. Volume is too low. Use rolling three-month periods.
Frequently asked questions
How is journey analysis different from funnel analysis?
They use the same numbers. Journey analysis adds time in stage and the reasons behind drop-out, and it includes post-sale stages such as onboarding and renewal.
Should we analyse by source or overall?
Overall first, to find the weak stage. Then by source, to see whether the weakness is general or limited to one channel.
What counts as a good stage conversion?
There is no universal benchmark. The useful comparison is your own stage against itself last quarter, and one source against another.
Can AI help with the analysis?
AI can summarise lost-deal notes and call transcripts to suggest common reasons, which saves reading time. The decision on what to change still belongs to a person. For mapping the path itself, see customer journey mapping.