Turning a GazeSite Report Into a Two-Week Action Plan

The most common fate of an audit report is to be read once, agreed with, and never opened again. This isn’t because the reports are bad. It’s because a report is a list of problems, and a list of problems is not a plan. Something has to convert one into the other, and that conversion step is where almost everyone stalls.

I built GazeSite to produce reports, so I’ve thought a lot about what happens after someone reads one. Each report covers six areas of a site and each finding comes with a prescribed remedy, so in principle the work is already specified. But “here are twenty-three things wrong with your site” produces a particular kind of paralysis. The reader’s instinct is to sort by severity and start at the top, and that instinct, reasonable as it sounds, is usually what kills the effort. The most severe finding is often also the most expensive to fix. You start on it, it drags, other work intrudes, and three weeks later the report is a tab you feel guilty about.

The fix is to stop planning by severity and start planning by time. Give yourself exactly two weeks. Not because two weeks is magic, but because it’s long enough to make real changes and short enough that you can actually protect it. A two-week plan forces the one question that matters: given ten working days, what subset of this report do I actually do? Everything else follows from answering that honestly.

Day one should produce no fixes at all. Spend it sorting the findings into three piles by cost, not severity. The first pile is anything you can fix in under an hour: a missing page title, an image that ships at full resolution, a link whose text says “click here,” a button with no accessible label. The second pile is work measured in an afternoon or a day: rewriting the hero copy so it says what the product does, restructuring a wall of text, fixing the mobile layout of a form. The third pile is anything that smells like a project, meaning it involves other people, a redesign, or infrastructure. Be ruthless about the third pile. Its purpose is to be excluded. A two-week plan that contains a project is a project plan with some errands stapled to it, and the errands will lose.

Days two and three are for the first pile, all of it, in one sitting if you can. This ordering looks backwards, since these are often the low-severity findings. But it’s right for two reasons. The practical one is that small mechanical fixes are pure execution, so you can’t stall on them, and finishing ten of them creates the momentum that carries you into the harder work. The strategic one is that these fixes compound quietly. Alt text, labels, titles, compressed images: none moves a needle alone, but together they change how the site reads to screen readers, to search engines, and to anyone on a slow phone.

The rest of week one goes to the single most severe finding that fits inside the two weeks. Note the qualifier. Not the most severe finding, the most severe one that fits. If the report’s top finding is that your information architecture confuses visitors, and fixing that properly means rethinking your navigation with three stakeholders, it goes to the third pile and you take the next one down. This feels like cowardice, and it’s the most important discipline in the whole exercise. A finding you fix beats a worse finding you gesture at.

Week two is for the second pile, and here you should cluster by page rather than by topic. If the report flags the pricing page for readability, conversion, and a technical issue, fix all three in one session. You’re already in that file, already holding that page in your head. Bouncing between pages to follow the report’s category order wastes the most expensive thing in this process, which is context.

Reserve the last day for the step everyone skips: verification. Run the scan again. This is the honest part. Some fixes won’t have taken. A remedy you thought you applied turns out to be half-applied, or applied on desktop but not mobile, which is exactly why GazeSite captures both. Seeing findings disappear from a fresh report is also, frankly, the reward, and you shouldn’t deny yourself the reward. It’s what makes you willing to do this again next quarter.

What about the third pile, the projects? Write them down somewhere real, with the finding attached as evidence, and then let them go for now. The evidence part matters. “We should redo the navigation” is an opinion that loses arguments. “The audit flagged this, here’s the finding, here’s what it costs us” is a case. The third pile isn’t work you’re refusing. It’s work you’re arming.

The underlying principle is that a report describes the site, but a plan has to describe you: your hours, your skills, your fortnight. Severity ranks the problems. Only time-boxing ranks the work. Two weeks, three piles, small things first, one big thing per week, verify at the end. It’s not sophisticated. That’s why it gets done.