Guide · traffic and sign-ups

Search traffic doubled but sign-ups did not rise: what happened? MochiStamp's last 7 days by language

Updated 2026-11-13 · kagentx

Short answer

From our data, the new clicks landed mainly on Persian and Arabic pages, while sign-ups only fell from 9 to 5, a difference of 4 people. In the last 7 days (2026-11-06 to 11-12) MochiStamp had 297 visitors (+123%), 124 search clicks (+130%), 5 sign-ups (−44%) and 16 generations; of the 124 clicks, fa pages had 49 and ar pages 20. Possible explanations include a sample that is too small, new visitors who only want the answer, and countries outside the product's main market; some have data, some are only assumptions, and we have no data on which page sign-ups came from, so we draw no conclusion.

How much did the last 7 days change?

Comparing the last 7 days (2026-11-06 to 11-12) with the previous 7 (10-30 to 11-05), MochiStamp went from 133 to 297 visitors (+123%) and from 54 to 124 search clicks (+130%), while sign-ups fell from 9 to 5 (−44%); generations went from 12 to 16, and paid orders were 0 in both weeks.

Metric10-30 to 11-0511-06 to 11-12Change
Visitors133297+123%
Search clicks54124+130%
Search impressions1,7553,248+85%
Sign-ups95−44%
Generate1216+33%
Share of engaged sessions48%40%−8 percentage points

Source: local dashboard, read 2026-11-13, GA4, Google Search Console and site statistics. The Search Console data for 11-12 was still incomplete when read (0 clicks), so the real search clicks are somewhat more than 124; the reason is in How many days is Search Console data behind? Why are the latest days low?.

First a bit of arithmetic: sign-ups going from 9 to 5 is a difference of 4 people. With single digits a week, random variation can produce a change like that, so it does not by itself prove conversion got worse.

Which languages did the new clicks come from?

Mostly from Persian (fa) and Arabic (ar) pages. Of the 124 clicks in the last 7 days, fa pages had 49 and ar pages 20, together 69 (56%); the week before these two languages had 22 (41% of 54). Of the 70 additional clicks, fa added 34 and ar added 13. MochiStamp started with Japanese LINE stickers, and the Japanese pages (those without a language prefix) had only 11 clicks.

Page language (by URL prefix)Clicks, previous weekClicks, last 7 daysImpressions, previous weekImpressions, last 7 days
fa Persian1549227774
ar Arabic720408836
Japanese (no prefix)611232383
zh-hant Traditional Chinese88128167
en English43238181
ru Russian134181
zh-hans Simplified Chinese0040
Other (pages outside the top 25 by impressions each day)1330548816
Total541241,7893,338

The dashboard keeps only the roughly 25 URLs with the most impressions each day and folds the rest into “other”; in the table above “other” holds 24% of clicks, whose language cannot be told. The page table's impressions total (3,338) is about 3% above the site's (3,248), and we have not verified why. So the table shows direction, not exact numbers; Impressions but no clicks: what to do with pages ranking 8 to 20 explains what the query and page tables each leave out.

Visitors point the same way: in GA4 by landing page, sessions on fa pages went from 16 to 50 and ar from 5 to 41; by country, US visitors went from 19 to 79, Iran from 15 to 51, Japan from 29 to 67, while Taiwan fell from 23 to 14. The top landing pages are MochiStamp's Persian reply sticker guide (29 sessions, 0 the week before), MochiStamp's sticker sheet splitter (24, against 22) and MochiStamp's Arabic WhatsApp sticker guide (23, against 3).

What could explain it, and which explanations have data behind them?

Here are the ones we can think of, sorted into “has data”, “partly supported” and “only an assumption”. We draw no conclusion, because the dashboard does not say which landing page the people who signed up came from.

Possible explanationIs there data?Basis, or what is missing
Sample too small, random variationSupported by dataSign-ups going from 9 to 5 is a difference of 4 people; only single digits a week.
The new traffic lands on informational pages and visitors only want the answerPartly supportedMany of the new landing pages are guide pages (the fa reply-sticker guide, the ar WhatsApp sticker guide); the share of engaged sessions fell from 48% to 40%. But the dashboard has no data on whether these visitors signed up.
New visitors want ready-made stickers, not to make their ownOnly an assumptionSome queries literally mean “how to download WhatsApp stickers” and “stickers to reply thank you”, which look more like searches for ready-made stickers; but we have no interviews or on-site behavior data, so we cannot prove their intent.
The new traffic comes from countries outside the product's main marketSupported by dataUS and Iranian visitors rose sharply; MochiStamp started with Japanese LINE stickers. But which country the people who signed up are in, there is no data.
There is friction at login or sign-upVery weakLogin opened 17 times (9 the week before), login clicked 14 (11), sign-up events 5 (7). The ratio from login click to sign-up fell from 64% to 36%, but the difference is only a few people; what the friction would be, there is no data.
A tracking problem (events not recorded)Largely ruled outSite statistics count 5 sign-ups and GA4 also has 5 sign-up events; the two agree.
Some of the new visitors are botsnot verifiedGA4 filters known bots, but we did not check this batch of visitors separately.

In this table none of the items can be taken as “the cause”. If we look into it later, the most direct step is to record sign-ups by landing page and country and compare again next time.

What does this mean for reading traffic data?

  1. 1. Doubled clicks do not mean doubled sign-ups

    If the part of the traffic that grew comes from pages or countries far from the product, sign-ups may not rise. That is what happened this week: the traffic grew mainly on Persian and Arabic pages, and sign-ups did not follow.

  2. 2. Break conversion down by page, country and source

    Break search clicks down first, then look at the engagement of those landing pages; the engagement of ChatGPT sources is compared in Does traffic from ChatGPT convert better than traffic from Google?.

  3. 3. With single-digit weekly numbers, do not judge by percentages

    “−44%” sounds alarming but it is 9 to 5. Stretch the period, or look at the cumulative figure.

  4. 4. Record which page each sign-up came from

    This is the item our dashboard lacks now; without it, every “why” can only be an assumption.

  5. 5. Do not look only at clicks; also check whether the language and page relate to the product

    How to write multilingual pages and link them: How do you write hreflang on a multilingual site? How does AI search treat it?; how Arabic and Persian pages perform in search and AI answers: Arabic and Persian pages in search and AI answers: what one small site saw. Where a new site's clicks come from: How long until a new site gets its first 100 search clicks? Four sites, real numbers.

What are the limits of these numbers?

  • One site and two 7-day periods; sign-ups are only 9 and 5.
  • The page table keeps only about 25 URLs a day and “other” holds 24% of clicks, whose language cannot be told; the page table's impressions total is about 3% above the site's, reason not verified.
  • The Search Console data for 11-12 is incomplete, so clicks for the last 7 days are low.
  • The dashboard does not say which landing page or country the people who signed up came from, so none of the explanations is confirmed.

How much of the search impressions and clicks come from the Arabic and Persian pages is in Do Arabic and Persian long-tail pages work? Data from MochiStamp's reply-sticker pages.

FAQ

Search traffic doubled but sign-ups did not rise: did conversion get worse?

Not necessarily. In our data sign-ups fell only from 9 to 5, a difference of 4 people, which random variation can produce at single digits a week. The new clicks also came mostly from language pages different from the product's main market.

How do you tell whether the new traffic is the “right people”?

Break it down by landing page, country and source and look at each group's engagement and sign-ups. Our dashboard still lacks “which page the sign-up came from”, so we could not judge this week.

Which countries did the new visitors come from?

On our site: the US from 19 to 79, Iran from 15 to 51, Japan from 29 to 67, Taiwan down from 23 to 14 (GA4 visitors, last 7 days against the week before).

Is it a tracking problem with sign-ups?

We checked: site statistics count 5 sign-ups and GA4 also has 5 sign-up events, so the two agree and it does not look like events were missed.

Should you remove pages that bring clicks but no sign-ups?

We would not: a week of data is too little, and visitors to these pages may come back later. Record sign-ups by landing page first and judge after a few weeks.

Sources

kagentx

Do AI answers
mention you?

An evaluation asks ChatGPT, Perplexity and Claude your buyers' questions (with web search on) and records who they recommend, which pages they cite and whether you are among them. Basic $1, results usually in 2–3 minutes.