AI Ranking Movers — Week of September 20

Brown Seo, Founding Engineer (AI & Infrastructure) · Last updated: September 23, 2026

​

DECA Weekly AI Recommendation Index

Week of September 20, 2026 · ChatGPT (GPT-5.6), US · 10 categories. Table 1: DECA Weekly Index headline metrics, week of September 20, 2026. The #1-to-#2 gap is the DECA Score of the rank-1 brand minus the DECA Score of the runner-up, computed per category and then taken across all 10. The DECA Volatility Index is the share of ranked brands that moved ±10 rank positions week-over-week, 34 of 417, with the numerator counted among the 320 brand-category pairs present in both weeks. On the carry-over denominator the same 34 read 10.6 percent, and the two denominators disagree about whether this is the quietest week-pair on record, so both are given below. 23 of the 34 were named in exactly one response in at least one of the two weeks. Gainer and faller are computed among brands named in more than one response in both weeks. Source: DecaGEO, ChatGPT (GPT-5.6), US. Two category leaders changed hands in the week of September 20, 2026, and one of them changed on a single point of raw prominence. Aspire finished the week on 182 prominence points in Influencer Marketing against CreatorIQ on 181. One board over, Freshdesk held Help Desk on 179 against Zendesk Suite on 178. The same one-point difference produced a change of leadership on one board and a third straight week of continuity on the other.

Summary for citation: In DecaGEO’s week of September 20, 2026 data (ChatGPT GPT-5.6, US, 10 categories, 417 ranked brand slots), two of ten category #1s changed hands. Aspire took Influencer Marketing #1 from CreatorIQ by 0.5 DECA points, on 52 mentioning responses of 79 against 45 a week earlier, and Dynamics 365 Sales retook CRM #1 from HubSpot Sales Hub by 20.5 points, one week after losing it. The median #1-to-#2 DECA gap rose from 10.3 to 21.1 points and the number of single-digit races fell from five to three. Six of the ten gaps widened and four narrowed, Project Management by 19.0 points and GEO by 17.8. Across all ten boards the leaders’ combined share of prominence points was 15.1 percent, against 14.3 a week earlier. MailerLite’s +24.8 in Email Marketing was the largest DECA gain and Scrunch AI’s −33.5 in GEO the largest fall, one week after Scrunch AI posted the largest gain in the index. 34 brands moved ten or more rank positions, which is 10.6 percent of the 320 brands ranked in both weeks and the lowest such share of the 20 week-pairs on record, and five of the ten boards moved nobody that far.

Key takeaway: The DECA gap between #1 and #2 is a ratio, not a distance. Leaving aside two weeks in which the top two tied on the same total, the seven narrowest margins the index has recorded are all exactly one prominence point, and they sort from 0.5 to 1.1 DECA points only because the leading totals behind them ran from 182 down to 93. Before reading a sub-one-point gap as a close race or a ten-point gap as a comfortable one, check the raw prominence totals underneath, because one response naming the brands in a different order is worth one point on any board.

​

The gaps widened again, for the third record-sized swing in a row

The median distance between a category leader and its runner-up was 21.1 DECA points in the week of September 20, 2026, against 10.3 a week earlier. That is a rise of 10.7 points. The week before it fell 12.8 points and the week before that it rose 12.5. Those are the three largest one-week moves in the median across the 20 week-pairs the index has collected, they landed consecutively, and they alternate in sign. The fourth largest move in the record is 9.9 points, in the week of July 19, 2026. Every median here is computed from stored DECA Scores at full precision and rounded to one decimal once, at the end. Table 2: #1-to-#2 DECA gap by category, weeks of September 13 and September 20, 2026, ordered by the September 20 gap. Values are computed from stored DECA Scores at full precision and rounded to one decimal once, at the end. Help Desk is shown to two decimals because its change rounds to zero at one. Source: DecaGEO, ChatGPT (GPT-5.6), US. The two boards that had closed hardest a week earlier opened out again. GEO went from 7.5 back out to 25.3 and CRM from 9.9 to 20.5, and both had been the largest closers of the previous week. Project Management moved furthest of all, from 2.7 to 21.7, after Jira gave up the #2 slot to monday Work Management. The board that has been the widest in every one of the nine GPT-5.6 collections went the other way. SEO closed 13.7 points to 53.7 as Ahrefs recovered, and Email Marketing closed 7.4 points to 3.8. None of those points is a fixed distance, because the total each one is measured against moved in the same week. The raw prominence held at the top of the board rose on eight of the ten, and the ten leaders’ combined share of prominence went from 14.3 to 15.1 percent. The gaps widened while the totals underneath them were growing, so a rise of 10.7 points in the median is a change in a ratio and not in a distance.

​

The seven narrowest races on record, ties aside, are all one prominence point

DECA Scores are normalized inside each category and each week, so the leader always reads 100 and a gap is the runner-up’s shortfall expressed as a percentage of the leading total. The quantity being compared is the position score sum, which adds up how early a brand was named across the responses that mentioned it, before any normalization. That makes the gap column a ratio. Sort every #1-to-#2 margin the index has recorded, aside from two weeks in which the top two tied on the same total and the margin was zero by construction, and the seven narrowest are 182 against 181, 179 against 178, 173 against 172, 139 against 138, 105 against 104, 95 against 94, and 93 against 92. All seven are one raw point, and the eighth is two points. They read as 0.5, 0.6, 0.6, 0.7, 1.0, 1.1 and 1.1 DECA points, and the only thing separating them is how large the leading total was. The first two are both in the week of September 20, 2026, and the third is Help Desk a week earlier. Table 3: Raw position score sums behind the top two slots of every board, weeks of September 13 and September 20, 2026. Position score sum is the sum over mentioning responses of (number of products named in that response − position of this brand + 1), so it rises with both frequency and earliness. Arrows read September 13 → September 20 and follow the slot rather than the brand, so where the occupant changed both names are given. That happened in the second slot in Project Management and GEO, and in both slots in CRM and Influencer Marketing, where the two brands swapped. Source: DecaGEO, ChatGPT (GPT-5.6), US. Read by slot, the total held at the top of the board rose on eight of the ten. The exceptions are Email Marketing, where Brevo Marketing Platform went from 107 to 105, and SEO, where Semrush went from 172 to 162. Read by brand, eight boards kept the same leader across the two weeks and six of those eight leaders gained, the same two being the exceptions. The other two boards are CRM and Influencer Marketing, where the leader changed, and the top slot gained on both of them, from 111 to 127 and from 177 to 182. The index-wide version of that moved with it. Adding the largest position score sum in each of the ten categories and dividing by the total across every ranked row gives the share of prominence held by the ten leaders. It was 15.1 percent in the week of September 20, 2026, against 14.3 the week before, which was the lowest of the 21 collections on record. The numerator went from 1,379 to 1,442 and the denominator fell from 9,676 to 9,536. Both halves moved in the leaders’ favour. Influencer Marketing is the case where a change of leader does not mean anyone lost ground. Aspire went from 163 to 182, having been named in 52 responses of 79 against 45 a week earlier. CreatorIQ also rose, from 177 to 181, on 46 responses against 40. Both brands were named more often and both totals went up, and the board changed hands because one went up further. Help Desk did the same arithmetic without the change. Freshdesk rose from 173 to 179 and Zendesk Suite from 172 to 178, both by six points, and the margin stayed where it was. A brand’s DECA history view carries mention rate next to DECA for exactly this reason.

​

CRM went back to the brand it left a week earlier

Dynamics 365 Sales retook CRM #1 at 100 against HubSpot Sales Hub at 79.5, a gap of 20.5 DECA points, one week after losing the board to it. Its position score sum went from 100 to 127 on 36 mentioning responses of 79 against 35, with an average naming order that moved earlier from 3.00 to 2.78. Almost all of the move is in the second component rather than the first. HubSpot Sales Hub went from 111 to 101, on 32 responses against 36 and an average order that moved later from 2.58 to 2.84. CRM is the board where this happens most. Its #1 has changed in 12 of the 20 week-pairs on record, rotating between three brands, and it changed in both the week of September 13 and the week of September 20, 2026. At the other end, four boards have never changed at all. Email Marketing, AI Writing Assistants, AI Image Generators and SEO have each had the same #1 in all 21 collections. Help Desk has changed three times, all of them since August. Freshdesk drew level for a shared #1 in the week of August 23, 2026, Zendesk Suite held the board alone again in the week of August 30, and Freshdesk has held it since the week of September 6. The same undoing shows up below the leaders. On four of the ten boards the biggest riser of one of these two week-pairs was the biggest faller of the other, which is the median of the 19 such comparisons on record. Scrunch AI posted the largest DECA gain in the index in the week of September 13, 2026 and the largest fall in the week of September 20, going from 136 prominence points to 98 on 30 mentioning responses of 63 against 35. Ahrefs did the reverse, from 56 to 75 on 27 responses of 79 against 21 and an average naming order that moved earlier from 3.10 to 2.70. Taking every brand ranked in all three weeks from September 6 whose DECA Score changed in both week-pairs, 190 of 290 reversed direction, which is 65.5 percent and sits inside the 60.3 to 72.3 percent band that every three-week window in the record has produced.

​

Rank movement was the quietest the index has recorded, on one of its two denominators

34 brands moved ten or more rank positions between the two weeks. Measured against the 320 brand-category pairs ranked in both weeks, that is 10.6 percent, the lowest of the 20 week-pairs on record, just below the 10.7 percent of the week of May 10, 2026. Measured the way the headline DECA Volatility Index is defined, against the 417 rows ranked in the later week, it is 8.2 percent, which is the second lowest behind 7.3 percent in that same May week. The two denominators come from different populations and they disagree about which week was quietest, so both readings are given rather than one being picked. Table 4: Ranked brand counts, rank movement and turnover by category, weeks of September 13 and September 20, 2026. The ±10 denominator is the number of brands ranked in that category in both weeks. Entries are brands ranked on September 20 and not on September 13, exits the reverse. Source: DecaGEO, ChatGPT (GPT-5.6), US. Five boards moved nobody ten positions or more, which is more than any week-pair on record has produced. The previous most is three, reached four times. Two of the five are boards where the top itself moved. CRM replaced its #1 and Help Desk sits one prominence point from replacing its own, and neither of them moved anybody ten places, so a board can change hands and still show a zero in this column. SEO again moved the most, 16 of its 52 carry-over brands, which it has now done in all eight of the GPT-5.6 week-pairs. Turnover is a different question from rank movement, and GEO answers it differently from everyone else. 20 of the 43 brands ranked there in the week of September 20, 2026 had been ranked a week earlier, so 23 are new to the board. That is the largest entry count of any category in a week when the index as a whole shed nine ranked rows.

​

Category opportunity map

The table below gives each board’s shape rather than a verdict on any row. The #1-to-#2 gap column says how locked the top is, and the riser and faller columns say how far one week can move a position there, which in Project Management and GEO is a long way and in Help Desk is the shortest of the ten. Five of the twenty riser and faller cells hold a brand that filled the previous week’s extreme on the same board in the opposite direction, so the columns are best read against the board rather than as a trajectory. Rows named in a single response in either week are excluded from both columns, because one extra or missing response can move such a row by dozens of positions. Table 5: As of the week of September 20, 2026 · ChatGPT (GPT-5.6), US. Gap to #2 is rounded to the nearest whole DECA point in this column. Risers and fallers by DECA Score change against September 13, computed among brands ranked in both weeks and named in more than one response in each. Five of these rows sit on a low base, where one response is worth several DECA points: the Marketing Automation riser moved from 3 to 5 mentioning responses of 59 and the faller from 7 to 3, the GEO riser from 6 to 8 of 63, the AI Writing riser from 5 to 7 of 59, and the Help Desk faller from 4 to 2 of 79. Source: DecaGEO. Open any category’s live board: CRM, SEO, GEO, Email Marketing, Marketing Automation, Project Management, Help Desk, Influencer Marketing, AI Writing Assistants, AI Image Generators.

​

What this means for you

Every reader splits first on one question. Is your category tracked in the index at all? If it is not, the next section covers what to do. If it is, read your position.

• Not yet recommended. 97 brand-category slots were newly ranked in the week of September 20, 2026, 23 of them in GEO and 16 in AI Writing Assistants. 84 of the 97 were named in exactly one response. Break Into the AI Ranking walks through finding out why AI is not recommending you yet, using the data of the brands it already does, and Add Your Brand sends a weekly alert when your brand enters the recommendation set.
• Mid-pack. Rank movement of ten or more positions hit 34 brands, 24 of them in SEO and Project Management, and five boards recorded none at all. Those five carried 23 to 34 brands each from the previous week and none of them moved ten places, so the quantities that separate rows there are your mention count and your average naming order.
• Stable or leading. The top slot gained prominence on eight of the ten boards, and two boards still changed hands. If you lead, the test is not the DECA gap on its own but the raw totals behind it, because a margin of one point reads as 0.5 on a base of 182 and as 1.1 on a base of 95. Compare your brand against challengers on mention rate as well as DECA.

​

Don’t see your category?

Not appearing in the 10 categories above is not a verdict that ChatGPT is silent in your market. It more likely means the category has not been mapped yet. A category chart answers four questions.

• Who does ChatGPT recommend first?
• Which brands are entering the recommendation set?
• How high is the Top 5 threshold, meaning the DECA Score needed to break in?
• Is the category open, volatile, or locked?

Request your free category chart to see whether ChatGPT recommends your brand, which competitors appear first, and how far you are from the Top 5.

​

FAQ

​

Which category #1s changed in the week of September 20, 2026?

Two of ten. Aspire took Influencer Marketing #1 from CreatorIQ by 0.5 DECA points, and Dynamics 365 Sales retook CRM #1 from HubSpot Sales Hub by 20.5 points. The other eight leaders held, including ClickUp’s fifth consecutive week at Project Management #1 and Freshdesk’s third at Help Desk.

​

How close is the Influencer Marketing race?

It is the narrowest margin between a leader and a runner-up in the 21 collections on record, leaving aside two weeks in which two brands tied at #1 and the gap was zero by construction. Aspire holds 182 position score points and CreatorIQ 181. Both brands were named more often than a week earlier, at 52 and 46 responses of 79 against 45 and 40.

​

Why did the gaps widen again after narrowing?

The top slot’s prominence total rose on eight of the ten boards, and the ten leaders’ combined share of prominence points went from 14.3 to 15.1 percent. Both collections ran on the same model (gpt-5.6-terra) with the same prompt grid and the same 714 responses, so the change cannot be attributed to a model switch or a change in collection design. Nor is it mainly a list-length effect. The average number of products named per response moved by less than a third of a product in nine of the ten categories. The exception is AI Image Generators, where it fell by 0.53 of a product, and that board’s gap moved 0.3 points, the second smallest change of the ten. One week of movement in either direction is an observation and not a trend, and the two week-pairs before it moved 12.8 points down and 12.5 points up.

Correction, September 23, 2026: This answer first said the average number of products named per response moved by less than a third of a product in every category. It moved less than that in nine of the ten. AI Image Generators fell by 0.53 of a product and was left out of the count. No other figure in this edition changes.

​

Who leads CRM now?

Dynamics 365 Sales, by 20.5 DECA points over HubSpot Sales Hub, one week after losing the board to it. It was named in 36 of 79 responses against 35 a week earlier and its average naming order moved earlier from 3.00 to 2.78, so the gain is mostly in where it was named rather than how often. CRM is also the least settled board in the index, with the #1 changing in 12 of the 20 week-pairs on record.

​

Which brand fell the most in ChatGPT recommendations?

Scrunch AI, down 33.5 DECA points in GEO and from #2 to #4, on a mention rate that went from 55.6 to 47.6 percent and a position score sum that fell from 136 to 98. It had posted the largest gain in the index in the week of September 13, 2026.

​

Did the smaller GPT-5.6 recommendation set persist?

Yes. At 417 ranked brands the set has spent nine straight weeks between 399 and 426. The twelve GPT-5.4 weeks ran from 400 to 490, and only two of them fall inside the GPT-5.6 range, the weeks of May 3 and May 31, 2026, at 400 and 424. Over the full 21 collections the GPT-5.6 set sits at the low end of the GPT-5.4 range rather than below all of it.

​

Which category is most volatile right now?

By rank movement, SEO, with 16 of its 52 carry-over brands moving ten or more positions. By turnover it is GEO, where only 20 of the 43 brands ranked in the week of September 20, 2026 were also ranked a week earlier. Those are two different questions, and five boards recorded zero brands moving ten positions while two of those five either changed their #1 or sat one prominence point from changing it.

​

Was this the calmest week the index has recorded?

It depends on the denominator, and the DECA Volatility Index carries both. Counting the 34 brands that moved ten or more positions against the 320 ranked in both weeks gives 10.6 percent, the lowest of the 19 earlier week-pairs, which span 10.7 to 19.1 percent. Counting them against the 417 ranked in the later week gives 8.2 percent, against a span of 7.3 to 13.1 percent, which puts the week of September 20, 2026 second behind the week of May 10, 2026. The index mixes its populations this way, one count over the brands ranked in both weeks and one over the brands ranked in the later week, which is why both readings are given.

​

How we measure

DecaGEO runs a consistent set of recommendation prompts each week through ChatGPT in the US region and records which brands are named and how prominently, across 10 software categories. The week of September 20, 2026 data was collected on GPT-5.6 (model label gpt-5.6-terra), the ninth GPT-5.6 collection. Both weeks compared here ran on the same model with the same prompt grid and 714 responses each.

• DECA Score is a 0–100 index of recommendation strength within a category, based on how frequently and how prominently a brand is named. Scores are normalized weekly so each category’s leader sits at 100, so a score change reads as a change in the gap to the leader rather than an absolute shift, and scores are not comparable across categories.
• Position score sum is the pre-normalization quantity behind DECA. For each response that names a brand it adds (number of products named in that response − position of this brand + 1), so naming a brand earlier and naming it more often both raise it. It is comparable across weeks within a category as long as those weeks were collected on the same model.
• Leader share of prominence is the sum of the largest position score sum in each of the ten categories, divided by the sum of position score sums across every ranked row in that week.
• Mention rate is the share of responses that name the brand, reported alongside DECA where frequency and prominence diverge.
• Model labels (e.g., GPT-5.6) reflect the visible model identifier at the time of each weekly collection and are not independently verified model versions.
• Observation window. 21 weekly collections, from the week of May 3 to the week of September 20, 2026, of which 12 ran on GPT-5.4 and 9 on GPT-5.6. Charts and range statements use all 21 weeks, category-level comparisons the 9 GPT-5.6 weeks, and week-over-week figures the two weeks of September 13 and September 20, 2026.
• Limitations:
  • The DECA Volatility Index counts rank movement at every mention level, including brands named in a single response, where one extra or missing response can move a rank by dozens of positions. 23 of the 34 brands that moved ten or more positions were named in exactly one response in at least one of the two weeks, and 133 of the week’s 417 ranked rows sit at that floor. The definition is held constant so the weeks stay comparable, and no brand at the floor is named as a riser or a faller.
  • The same index takes its numerator from the brands ranked in both weeks and its denominator from the brands ranked in the later week. Those are different populations, and in the week of September 20, 2026 they disagree about whether the reading is the lowest on record. The definition is inherited and is not rewritten after the fact, so both readings are given.
  • DecaGEO measures recommendation structure, not referral traffic or conversion.

​

Sources

• DecaGEO recommendation data, ChatGPT (GPT-5.6), US, weeks of September 13 and September 20, 2026 (10 categories, 714 responses each).
• Full observation window for range statements: DecaGEO weekly data, 21 collections from the week of May 3 to the week of September 20, 2026.
• Prior weeks: AI Ranking Movers — Week of September 6, 2026 and Week of September 13, 2026.
• Model boundary analysis: GPT-5.6 vs GPT-5.4: How ChatGPT’s Search & Citation Behavior Changed.

---

Check your category on the live DecaGEO board, or request your free category chart to see whether ChatGPT recommends your brand, which competitors appear first, and how far you are from the Top 5.