How Scaup gets ranking data, how to explain it to a customer when our number does not match their own Google search, and what we changed recently to make this less of a problem.
This comes up often. Almost always the answer is: the data is correct, but it is measuring something different from what the customer just did.
Every position you see in a Scaup site report comes from one of two sources. You should know both.
For the top keywords in the report, we run a real Google search from the customer's primary country and read off where their site lands. This is the number we trust as "current rank." It is fresh, country-specific, and reproducible.
For everything else, we use the average position from the customer's own Search Console account over the last 28 days. This is good for trend and movement, but it is an average across many situations.
Even the live check will not always match what the customer sees. Here are the six reasons. Pick the one that fits the situation:
Where they are, what device they use, what they searched before, whether they are signed in. All of it shifts the results. Two people searching the same term from two coffee shops in the same city can see different rankings.
If the site ranked #1 on 10 impressions and #20 on 10 impressions, GSC shows a low average. A manual search shows one current position. Neither is "wrong". They answer different questions.
Hebrew queries return very different results from Tel Aviv vs Haifa vs outside Israel. We filter our data to the customer's primary country, but a customer searching while traveling, or using a VPN, will see different results.
In GSC, impressions for a long phrase can be attributed to the short root query inside it. The site might rank well for the long phrase but poorly for the short one, and GSC mixes them.
A very recent change in rank will not appear yet. If a customer says "but I just dropped yesterday," GSC will not show it for a couple of days.
If a keyword only had one impression, even one lucky placement at #1 makes the GSC average look like #1. We filter out anything below 5 impressions, but edge cases still happen.
"Good question. Let me explain what you are looking at. The position in the report is not your personal Google result. It is a live check we ran from [country], without personalization. So your own search, from your phone, signed in, in your specific city, will almost always look a little different. That does not mean either number is wrong. They are measuring two different things."
"That is expected. Google shows different results to every person based on their location, history, and device. The number we show is the closest thing to a neutral reference. It is what Google returns for a clean search from your country. We work to improve that number, and over time you will see your personal results follow."
"That is also expected. Search Console gives you the average position across all the times Google showed your site in the last 28 days. Our live number is a snapshot from today. The two will not match unless the site has held the exact same position for a full month, which is rare."
"That can happen when Search Console has a small number of impressions for a very specific phrasing where the site really did rank high, but only once or twice. We filter most of those out, but if you see this, send it to us and we'll review the exact keyword."
Up until recently, the report relied only on Search Console's 28-day average. That is what caused most of the "your ranking is wrong" complaints. We just shipped three changes that make the number the customer sees much closer to reality.
For the top keywords in the report, we now run an actual Google search from the customer's primary country (Israel, US, UK, etc.) and read off the live position. That number is reproducible and current, not an average. When the report says "You currently rank #X," that is what came back from Google today.
Before, GSC averages mixed traffic from every country Google logged. A site that ranks #3 in Israel but #45 in the US could show an average of #24, which is useless. Now we filter GSC by the site's primary country before doing anything with it. The average reflects the audience the customer actually cares about.
If a keyword only had 1 or 2 impressions, GSC's "average position" was statistically meaningless. One stray placement at #1 would skew the average to #1. We now drop anything with fewer than 5 impressions before showing it in the report. This kills most of the "but I can't find myself anywhere" complaints.
השאלה שתקבלו מהלקוח: "חיפשתי את המילה בגוגל ואני לא באמת במקום 4. הנתון שלכם שגוי?"
הדירוג שמופיע ב-Scaup הוא בדיקה חיה שאנחנו מבצעים בגוגל מתוך המדינה של הלקוח, ללא התאמה אישית. החיפוש שהלקוח עשה בעצמו מותאם אליו אישית – למיקום שלו, למכשיר, להיסטוריית החיפושים ולחשבון שלו – ולכן הוא כמעט תמיד ייראה קצת אחרת. אף אחד מהמספרים לא "שגוי", הם פשוט מודדים שני דברים שונים.