Your Search Console, turned into this week’s list of changes
Search Console tells you where you appear. SEOLift joins it to the reading of your site and turns it into a short list for the week: the exact change, a forecast as a range, and a measurement once you have made it.
Seven goals, not a dashboard
Pick a goal, get the list of pages that lead to it.
Search Console already holds the searches, the impressions and the positions for every page you have. On its own it does not tell you which page to work on this Thursday. SEOLift joins that data to the reading of your site and sorts every opportunity into the goal it serves. You pick the outcome you want, and you get the pages that lead to it.
There are seven goals. Each is a fixed set of rules over your own rows, and each shows the estimated gain and how many of its changes need no new backlinks.
- Page two to page one: pages sitting at 11 that are one change from page one.
- Top ten to top three: page-one results nobody clicks, and what holds them back.
- More clicks, same ranking: same position, more clicks. See which title is losing.
- Win back lost traffic: pages that used to bring more, and why they stopped.
- One page per search: two of your pages fight for one search. Know which to keep.
- Answer the questions people ask: questions people type that your page never asks.
- Pages worth writing: searches you already show for with no page behind them.
Five for this week
Five things for this week. The rest can wait.
Above the goals sits This week: five changes at most, best first. Each has a rough effort, from fifteen minutes to a day. A line under them says how many more are waiting, and why they can wait.
Before any list, a short note says what the data can see. Google withholds rare searches, so the rows never explain all of your impressions. The note shows how much they do explain, how much is your own brand, and how much lands on addresses that are not pages, such as redirects. When the rows see less than 60% of the demand, it says the lists are partial.
What would prove it wrong
Every recommendation says what would prove it wrong.
Every opportunity reads the same way, in the same order: what we found, which data says so, why it matters, the exact change, the expected result, how sure we are, and what would prove us wrong. The change comes from a short vocabulary: rewrite the title and H1, add internal links, ask the question in a heading, update the content, redirect or merge, write a new page, and a few more.
The expected result is a range of clicks, never a single number, and code computes it, not a model. Where the change is text, a model may draft a title or a description, and code checks the draft: its length, its width on the result page, the search it has to cover, and that every number in it already appears on your page. A draft that fails twice is shown as a failure, with the reason.
The risks are named on the card, before you touch anything. That is not hedging. A tool that only tells you when it is right gives you no reason to doubt it on the day it is wrong.
Measured against pages nobody touched
After each change you know if it worked, against pages nobody touched.
Nothing changes on your site by itself. You approve an opportunity, dismiss it with a one-line reason, or mark it done. Then the module goes looking for the change instead of believing it happened. A measurement starts only once a reading has seen the change on the page, and the clock starts on the day it was seen.
Your page is compared with similar pages of the same template that nobody touched: similar traffic, the same band of positions. What you see is the effect of your edit, not the season or the news. When there are not enough such pages, the rest of the site stands in, and the verdict can be probable at most.
The verdict is one of a few plain words: confirmed, probable, no effect, negative when the change cost traffic, or too early. When the numbers cannot decide, it says inconclusive. The unflattering verdicts are the point.
When it says nothing
No data, no advice. It says so plainly.
An empty list is a real answer. It means no page on the site is close enough to winning to be worth your afternoon, and the screen says so in those words.
No rule speaks with less than ninety days of Search Console data. A page that does not exist yet gets no click forecast, because there is no history to count from. A page whose competitors outrank it on links is told so, with the numbers, and gets no forecast either, because no change on the page closes that gap. A search nobody makes any more is left alone.
Both stay visible at the bottom of the goals screen, under what we looked at and are not recommending, so you know they were seen.
Refresh before the clicks die
Losing clicks? Know which section to refresh, not the whole page.
Pages losing clicks against their own history of ninety days earlier sit under Win back lost traffic. When the cause is technical, such as a canonical, a noindex or a slow response, it stays in the technical reading. When the cause is the content, the opportunity names the searches that fell and the one section of the page that answers them.
The draft is that section. Update the figure, the price or the screenshot in it, then send the one-line IndexNow request that comes with the draft. Fifteen minutes to a day, not a rewrite. On Agency, with server logs uploaded, the check after your change also says whether Googlebot and Bingbot fetched the page again, and when.
What it never does
It does not invent numbers. Every click figure comes from your own rows and is computed by code, and a draft with a number your page does not contain does not pass. It publishes nothing and changes nothing on your site: you make the change, it measures it. It never promises a ranking. What it gives you is smaller and checkable: a list for the week, the reason behind each item, and an honest answer afterwards.
SEOLift comes with the Business and Agency plans and stays off for every site until you switch it on, because reading a Search Console property is not something to start behind your back.