What a conversion rate actually tells you

A client sends us a screenshot most weeks: "our conversion rate is 1.2%, is that bad?" The honest answer is that the number alone tells you almost nothing. A 1.2% rate on a page selling a $40,000 piece of equipment to a cold audience is probably excellent. The same 1.2% on a page collecting email addresses for a free newsletter is a real problem. Conversion rate is a ratio, not a verdict — and ratios only mean something next to a baseline for that specific offer, that specific traffic, and that specific action.

The number is standing in for three different things

When a conversion rate is low, it is because of exactly one of three causes, and each one needs a different fix. Confuse them and you will spend a month redesigning a button when the real problem was who you were showing the page to.

1. The traffic is wrong

If people arriving at the page were never going to want what you sell — the wrong audience, the wrong intent, the wrong stage of deciding — no amount of page polish moves the number. This is the most common cause we find, and the easiest to miss, because the page itself often looks fine. The tell is a high bounce rate combined with very short time on page: people are leaving fast because the page answered their question in the first three seconds, and the answer was "not for me."

2. The page is unclear

The right people arrived, stayed a while, and still did not act. This is a clarity problem: they could not tell what you were offering, what it cost, or what happened next if they clicked. The tell is decent time on page with a lot of scroll-up behaviour — people re-reading, looking for information that was not there or was buried.

3. The offer does not fit

People understood exactly what was on offer and decided against it — on price, on trust, or because a competitor's page answered a doubt yours did not. This is the hardest one to fix with a page edit alone, because the problem often sits upstream of the page: pricing, guarantee, proof, or positioning.

How to tell which one you have

Before touching the page, look at three numbers together: bounce rate, average time on page, and scroll depth, segmented by traffic source. A page with a 70% bounce rate from paid search but a 30% bounce rate from an email list is not a page problem — it is a targeting problem specific to that ad. A page with low bounce but low conversion across every source is much more likely a clarity or offer problem, and that is where session recordings earn their keep: watch ten real visits and you will usually see the sticking point within the first three.

This is also why we build analytics into a site at launch rather than bolting it on later — segmented data from day one means this diagnosis takes an afternoon instead of a guess.

What actually moves the number, in order

• Fix the headline and first screen before anything else. If someone cannot tell what you do and for whom within three seconds, everything below it is wasted work.
• Remove friction before adding persuasion. A form with two fewer fields usually beats a better headline. Count every field, every required click, every page load between interest and action.
• Answer the specific objection, not a generic one. "Is this secure?" and "will this work for a business my size?" need different proof. Testimonials that do not address the actual hesitation do not help.
• Only then, test variations. A/B testing a button colour before fixing a genuine clarity problem is optimising noise. Test after the obvious problems are gone, not instead of fixing them.

The baseline that actually matters

Forget industry-average benchmarks — they average together offers, audiences and price points that have nothing to do with yours. The only useful baseline is your own page, last month, against this month, segmented the same way both times. Everything else is a distraction dressed up as a statistic.