Trust Signals, Proof and Case Studies

The four numbers worth tracking

You don't need a complex analytics setup. Four numbers per quarter, written on a single page, are enough to keep the proof library honest. The four are: enquiries (or sales) per hundred unique visitors, time spent on the offer page (a rough proxy for how engaged visitors are with the proof there), the proportion of visits that include a click to the case studies page (a sign that proof is being sought) and click-through rate from the offer page to the call to action. The website analytics eBook in this category covers the practical setup. For proof measurement specifically, these four are the ones that move when proof is doing or not doing work.

The two experiment shapes

Experiment one: the before-and-after

The simplest. Note the four numbers for the four weeks before the change. Make one change to the page (add a quote, replace a case study, move a guarantee). Note the four numbers for the four weeks after. If the conversion rate moved by more than fifteen percent in either direction, the change probably did something - keep the change if it was positive, reverse it if negative. If the conversion rate moved by less than fifteen percent, the change probably did nothing notable - the noise in small business website analytics is usually larger than that. Move on to the next change.

Limitations: small businesses don't have enough traffic for proper statistical confidence on most changes. The before-and-after method is rough. It's still worth doing because it builds the habit of paying attention, and because the changes that move the numbers a lot (the right guarantee in the right place, the right case study near the right tier) tend to move them visibly even on small samples.

Experiment two: the qualitative read

For changes too subtle for the numbers to detect cleanly. Show the page (before and after the change) to three people outside your business. Ask them three questions about each version. What did you remember from this page? Which version of the page would make you more likely to enquire? Why? Note the answers verbatim. The qualitative read often catches things the analytics miss, particularly when proof is improving the felt experience without the visitor being clear it's the proof making the difference.

Three signals that proof is doing work

First: the headline conversion rate climbs after a placement change. Second: the click-through from the offer page to the call to action improves. Third: incoming enquiries start mentioning specific case studies or quotes ('I read about Tom Hughes' situation and we're in the same boat'). The third signal is the most valuable - it tells you that proof is reaching the visitor and changing their decision before they email.

Three signals that proof is not doing work

First: the conversion rate is flat after months of adding proof. The proof is being added but not in the placements where decisions happen, or it's not addressing the worries actually being felt. Second: the case studies page gets traffic but conversion doesn't lift. The visitors are reading proof but the proof is not reaching the people about to buy - it's reaching the curious, not the decisive. Third: customer feedback on enquiry forms says variations of 'I needed to know more before deciding'. The proof on the page didn't fill the gaps it was supposed to fill.

When to leave a change in place

Two reasons. The numbers visibly improved (the change is doing work). Or the numbers didn't move but the change closes a known gap in the placement map and you're confident it'll matter for the visitor population that hasn't yet arrived (the gap-closing change). The second case is common for small business sites where the experimental sample is too small to confirm in four weeks. If a change is in the right direction by the placement map, leave it in even if the numbers haven't yet confirmed it - you're betting on the structural improvement, not the short-term measurement.

When to remove a change

Two reasons. The numbers visibly worsened (rare but does happen, usually when an added piece of proof is clearly the wrong fit - a corporate testimonial on a tradesperson page). Or the change is duplicating proof that already exists elsewhere (the same quote in three places). Removing duplicate proof rarely hurts the headline number and usually tightens the page; it's a low-risk move when in doubt.

The quarterly review

An hour a quarter, on a Thursday afternoon. Open the proof inventory and the four numbers from the last three months. Ask three questions. Which placements have moved the numbers since the last review? (Keep them.) Which placements haven't moved anything in two consecutive quarters? (Consider replacing or removing.) Which new proof has come in this quarter that should be placed somewhere? (Add it to the map.) Then act on the answers - retire one or two pieces of proof that aren't carrying their weight, place one or two new pieces, plan one before-and-after experiment for the next quarter. One hour. Once a quarter. Four times a year.

Worked example: Sam's first three quarterly reviews

Quarter one (after building the inventory and placing per the map). Conversion: twenty-two percent of enquiries became customers, up from sixteen percent the previous quarter. Time on offer page: up from forty seconds to one minute twenty. CTA click rate from offer page: up from eight to fourteen percent. Three of the changes moved numbers visibly. One (a credential added to the home page footer) moved nothing measurable but stays in - the placement map says it should be there.

Quarter two. Conversion now thirty-six percent. The Tom Hughes case study below the offer page menu is being mentioned in roughly half of all enquiries ('we're in the same boat as Tom'). Sam writes up two more case studies (one electrician, one carpenter) using the same structure, places them on the relevant offer pages. Replaces a generic 'satisfaction guaranteed' line at the checkout with the specific 'tax return filed on time or your next month is free' guarantee from chapter four.

Quarter three. Conversion holds at thirty-five percent. Sam runs the qualitative read on three friends-of-friends - all three remember the Tom case study from the offer page; only one remembers the home page above-fold quote. Sam swaps the home page quote for a shorter, sharper one. Conversion holds. The proof library is now stable and the changes are smaller, more incremental, more about freshness than structural correction.

When to bring in help

If the analytics setup is the bottleneck (you can't reliably measure the four numbers), an analytics specialist can set up a basic dashboard in a few hours. The brief is narrow - 'I want to see these four numbers per page per month, that's it, no further reporting needed'. If the proof writing is the bottleneck, the same freelance copywriter relationship from the storytelling eBook applies - they write up the customer interviews using your structure, you keep the placement and measurement work. Most small businesses can run the measurement practice themselves once the dashboard is set up.

What to do this week

Set up a single sheet with the four numbers, written down for the last four weeks. Even if your analytics are imperfect, an approximate number you write down beats a precise number you never look at. Add one calendar entry for ninety days from now to run the first quarterly review. Pick the next placement change you'll make using the inventory from chapter five. Note the before numbers; make the change; note the after numbers in four weeks. The habit is the asset, not any single experiment.

Now turn the project into a sustainable practice. The next chapter, 'Building and Maintaining a Proof Library Over a Year', covers the documents, the gathering rhythm and the small habits that keep the proof fresh and ready to use rather than calcified into a static set.