Scrunch alternatives for a small team or for an agency: one published line answers both questions in opposite directions, and it is five licences inside one brand workspace

Read on 21 October 2026, Scrunch publishes one buyable plan, Core, and two numbers on it decide this page. Five user licences and one brand workspace. Five licences is the most generous seat count on any entry plan in this set that puts a number on seats at all. One brand workspace means a second client requires a quote. So the same plan is a good small team product and an unusable agency product, and which of those you are decides everything else.

The rest of Core is counted in ones. One country, one persona, one language, five competitors, three AI search trend topics, twenty five site map pages, five site audits a month, one page optimisation a month. A single team selling one thing in one market in one language fits that shape exactly. An agency does not fit it at all, and neither does a brand selling in two countries.

We sell a tracker here, so our own entry terms are in the table with everybody else's, including the seat line where Scrunch and three others are ahead of us.

The criterion: how many of everything on the plan you can buy

One criterion, and on a counted plan it is the only one that matters: on the cheapest plan a buyer can purchase without a call, how many of each thing. Read 21 October 2026, two agents each, bodies saved.

cheapest buyable plan, read 21 Oct 2026	client containers	seats	countries or markets	competitors
Scrunch Core	1 brand workspace	5 user licences	1 country, 1 persona, 1 language	5
Otterly.AI Lite	1 workspace, unlimited from the second tier up	unlimited team members	unlimited countries and languages	unlimited brands and competitors
Peec AI Starter	1 project	unlimited users	multi country from its third tier	published per tier in its matrix
Depra AI Starter	1 project	not stated on the pricing page	any of 40 markets, on every plan	3
Rankscale, lowest monthly plan	10 brand dashboard slots	no per seat charge	all regions, on every plan	published per plan in its matrix
AI Knows Us, ours	a client brand, set with a calculator for the agency plan	one login with full access, plus an export	published per plan	published per plan

On that criterion Otterly.AI is first for a small team and the margin is large: unlimited team members, unlimited brands and competitors, and unlimited countries and languages on its entry tier, with the prompt count as the only real limit. Rankscale is first for an agency starting out, because ten brand dashboard slots on its lowest monthly plan is nine more clients than Core allows and the slots are purchasable individually inside the app.

We are last on seats and it is the clearest loss in this table. Four of the five vendors above publish unlimited seats, unlimited users or no per seat charge. Every plan of ours publishes one login with full access and an exportable tracking log for the rest of the team, and our published reasoning is that the cost of this work is asking the assistants rather than headcount. For a five person marketing team that wants five people in a dashboard, that reasoning is ours and the product they want is theirs.

The co-naming measurement that an agency should take to a pitch

Our standing dataset is 1,260 readings from ChatGPT and Gemini across the United States, the United Kingdom and India; 718 name at least one tracked vendor, counted once per reading in which the name appears in the answer prose. Scrunch is named in 18 of those 718, 2.5%. Small, and the pattern inside it is sharp.

named in the same reading as Scrunch	share of the 18 Scrunch readings	share of all 718	ratio
Peec	55.6%	25.2%	2.20
Profound	50.0%	28.3%	1.77
Semrush, an Adobe company	50.0%	55.3%	0.90
Otterly	44.4%	29.9%	1.48
Ahrefs	16.7%	31.1%	0.54

Peec at 2.20 times its own base rate is the highest co-naming ratio between any two vendors in this set. An assistant that names Scrunch names Peec more than half the time, and names Ahrefs about half as often as it otherwise would. Eighteen readings is a base where one reading moves a figure, so this is published with its size attached. The direction is consistent with the wider pattern: the specialists travel together and the two big search suites sit outside that group. Our own name appeared in none of the 1,260.

In the small number of framed readings the base shrinks further: two readings name Scrunch in a money question, two in an agency question, one in an enterprise question and none in a reporting question, and none of those produced a numbered shortlist. For an agency pitching against a Scrunch incumbency, the useful consequence is that the client has probably not been handed a ranked comparison, and a comparison is the thing you can bring.

Which alternative fits which side of the Core line

• You are one team and you ran out of prompts. You probably have not. Core publishes 125 unique prompts, the largest entry prompt count in this set, with a separate ceiling of 5,000 responses a month. Check which of the two you actually hit, because the answer changes the vendor.
• You are one team and you need a second country. Core publishes one country, one persona and one language. Depra publishes 40 markets on every plan and Otterly publishes unlimited countries and languages on its entry tier.
• You are an agency and you need a second client. Rankscale publishes 10, 25, 50 and 100 brand dashboard slots across four monthly plans. Otterly publishes unlimited workspaces from the second of its four tiers. Depra publishes 5 pooled client projects with prompts movable between them.
• You are an agency and the report needs your name on it. Depra publishes logo, colour and footer control on client reports and Rankscale publishes white label options from its third plan. Scrunch does not state branding on its pricing page, and neither do we.

The two numbers Scrunch publishes that nobody else does

Two Core lines are worth copying into any comparison because they describe limits other vendors leave silent. The first is query fan out, published as Limited on Core and a tick on Enterprise. Query fan out is the assistant rewriting your question into several of its own before answering, and it is the mechanism by which a brand gets missed on a question it thought it was watching. A vendor that puts a limit on it is telling you something real about what the cheap plan sees.

The second is the site map page count, at 25 on Core. That is how much of your own site the audit will look at, and twenty five pages is a small site. For a business with four hundred pages, the audit on Core covers six per cent of it, which is a number worth knowing before the first report arrives rather than after.

Working out which published limit binds first, rather than which headline number is biggest, is the whole skill here, and the unit arithmetic that decides which prompt tracker you can use sets it out across more vendors. That page is one of the 1,068 of ours, 53.7% of the 1,987 in our inventory, that no stored body on this site links to.

Questions about Core, from both kinds of buyer

How many people can use Scrunch Core?

Five user licences, with custom counts above it. That is a real number where several rivals simply say unlimited, and for a team of five it is exactly enough.

Can an agency run two clients on Core?

Not on the published terms. Core publishes one brand workspace and custom workspaces on Enterprise, so a second client is a conversation rather than a setting. Rankscale, Otterly and Depra all publish multi client counts on plans you can buy.

Which engines does Core cover?

Four: ChatGPT, Perplexity, Google AI Overviews and Copilot. Its Enterprise tier publishes nine, adding Claude, Gemini, Meta AI, Google AI Mode and Grok. So five surfaces, including two of the largest, are a tier rather than an add on.

Did the Scrunch domain change?

scrunchai.com now redirects to scrunch.com, verified today under both agents, so an old bookmark or an old citation still resolves. Worth knowing if you are checking a comparison written a few months ago.

What is the one question to ask before buying any counted plan?

Which of its numbers you will hit in month two. On Core the candidates are one country, one language, five competitors, twenty five audited pages and 5,000 responses, and the prompt count of 125 is almost never the one that binds. A plan is only as big as its smallest relevant number.