Reading a comparison scorecard and prioritising by effort and impact

What a scorecard is, and what it is not

A comparison scorecard puts your site and your rivals' sites side by side across the same categories, each score backed by evidence. In an Intel loop report those categories are first impression, navigation and discovery, product pages, cart and checkout, trust signals, mobile experience, speed and basic SEO, and promotions and retention.

It is a picture of what a visitor can see. It is not a measure of sales, traffic or conversion, because none of those are visible from outside. Read every finding as an observation and a hypothesis: 'this looks like it costs us sales' rather than 'this costs us sales'. You can then test the ones that matter.

Two buckets: quick fixes and bigger opportunities

Start by splitting the findings into two groups. Most comparisons produce both.

Quick fixes

These are changes you or your developer can usually make without a project plan. Typical examples:

• a vague or buried value proposition on the homepage
• thin product pages: short copy, few images, no reviews shown, unclear delivery or returns information
• trust signals your rivals show and you do not, such as guarantees, a review widget or clear payment information
• no email sign-up or incentive where rivals have one
• slow-loading images or unoptimised pages
• weak page titles and meta descriptions
• checkout friction: forced account creation, delivery costs that appear late, limited payment options
• cluttered navigation or a site search that returns poor results

Bigger opportunities

These need design, development or a change in how the business works. They often matter more, but they take planning:

• a poor mobile experience that needs real design work
• no loyalty, referral or subscription offer
• no bundles, upsell or cross-sell logic
• thin content or SEO structure compared with rivals
• an outdated theme or platform that limits what you can do
• no sign of ongoing testing of offers or landing pages
• weak post-purchase and retention emails

The four patterns, and how to respond

Before reading individual findings, look at the shape of the whole scorecard. It usually falls into one of four patterns.

You are behind on most categories

The easiest pattern to act on: the job is choosing. Lead with the biggest gaps and pick around five quick fixes to do first.

Mixed results

The most common pattern, and usually the most credible. Note where you win, because those strengths are worth protecting and saying louder. Then look at the categories where a rival is pulling clearly ahead and ask why.

You are ahead

If the basics are covered, the comparison has done its job by confirming it. The next gains are less likely to come from the site itself and more likely to come from scale: advertising, retention and testing. The channel comparison is the next thing to read.

One competitor dominates

When one rival scores well across the board, use it as your benchmark. Go category by category and write down what the leader does differently, then decide which differences matter for your margins and customers. In the sample report, run on a constructed firm, Crowthorne Legal leads with The Anxious First-Timer: 81 against Hadley & Whitmore's 58. Crowthorne puts its process on the homepage in five steps. Hadley & Whitmore has no “How we work” page, shows its SRA registration only in the footer and uses a stock courthouse photo in the hero. Copy nothing; write down the difference and decide whether it matters to your buyer.

The effort and impact worksheet

Next, rate each finding on two scales: effort (low, medium, high) and impact (low, medium, high). Keep the ratings rough. The point is to sort, not to forecast.

A simple worksheet has five columns:

1. Finding, in one sentence with the evidence: 'Rival B shows delivery cost on the product page; we show it at the last checkout step.'
2. Effort: low, medium or high, based on who has to do the work and how much of the site it touches.
3. Impact: low, medium or high, based on how many visitors it affects and how close it sits to the moment of buying.
4. Quick win? Yes if effort is low and impact is medium or high.
5. Agreed? Whether the people who run the business agree the finding is real.

Sort by the quick-win column first, then by impact. The top of the list is your first batch of work. High-effort, high-impact items go into a separate plan with their own budget. Low-impact items, whatever the effort, can usually wait.

If you then turn the sorted list into a 30/60/90 plan, keep the same ratings and the same evidence alongside each item, so you can check progress against them later.

Ask what you already suspect

Before you finalise the list, ask the people closest to the customers what they already suspect is wrong. The customer service lead, the person who answers the phone, the developer who gets the bug reports.

Where a finding matches something they already believed, give it extra weight. Agreement between outside evidence and inside experience is the strongest signal you will get without running a test. Where a finding surprises everyone, do not dismiss it, but check the evidence again before you spend money on it.

This step also gets people on side. A change that someone has been asking for, now backed by evidence from rivals' sites, is much easier to schedule.

The swap-the-brand-name test

'Improve your calls to action' or 'post more often' fits any business, so it helps none. Test every finding: swap in another company's name. If the finding still makes sense, it is too vague to act on.

A finding that passes the test names something specific: the exact wording on your homepage, the step in your checkout where delivery appears, the rival that shows a returns promise you do not. Rewrite any finding that fails until it passes, or drop it.

Re-score after you act

Once a batch of changes is live, score the affected categories again using the same standard and the same rivals. The scorecard is most valuable as a before-and-after record: it shows whether you closed a gap, and it gives you evidence when someone asks whether the work was worth doing.