ICP search and content benchmark · 15 October 2026

Merchants searched 31 times. Outcraft appeared 0.

We searched the way our ecommerce buyers search, then measured the five companies that kept showing up instead of us.

One page a company has publishedA page that ranks on Google
Scroll or press ↓

01Executive summary

We are absent where buyers decide, and our core story is no longer unique.

0/ 31
natural ICP searches where outcraft.ai appeared
0vs 60
our Domain Rating vs the competitor median
0%
of our organic visits come from four off-ICP posts
0rivals
already publish a page for all four of our “moments”; we have none live
1

Five real competitors, not nineteen

Of 110 vendors that surfaced, 29 sell AI calling and 5 compete directly: Consio, Ringly.io, Callsy AI, Loop Voice and Thoughtly. Three of them are not in the Brand Book.

2

They win at the decision stage

Shopify App Store listings, compare pages and directory profiles put Consio in 6 of 11 decision searches. Ringly.io is the only one with an organic engine (≈10,700 visits a month).

3

Discovery is open for whoever moves first

In 9 of 10 symptom-led searches no AI-calling vendor appears at all. Our planned pillar pages target exactly this gap.

02Our ICP and how they search

Merchants search their symptom, never our category.

ICP (Brand Book v4)

Ecommerce brands on Shopify, US and Europe

Who
DTC brands, online retailers, marketplaces
Size
≈$1M–$50M revenue, running paid traffic
AOV
$100+, strongest at $200+
Lead segment
Subscription and replenishment brands (payments pillar)
Buyers
Ecommerce Manager, Head of Growth, CMO, Founder
Not a fit
Sub-$50 AOV fully self-serve stores

What merchants type vs what we call ourselves

US monthly searches; our positioning terms in blue

The buying journey, as merchants search it

1

A bad number

“abandoned cart emails not converting anymore”

Blames the email, not the channel. Searches in Google, short and lowercase.

Owned by: Shopify Community, email-recovery apps
2

An approach, and a fear

“is calling customers who abandon cart worth it or is that annoying”

Curious about the phone, scared of brand damage. Appends “reddit” to hear from operators.

Owned by: GrowthSuite (“use discounts instead”), Ringly.io, Consio
3

A vendor shortlist

“ai that calls abandoned carts shopify”

Shops the way Shopify merchants shop: app listings and “apps like X”.

Owned by: Shopify App Store listings
4

Risk and price

“is it legal to call customers who abandoned checkout in the us”

Legal exposure and price are the last blockers before a demo. Long, contextual prompts in ChatGPT.

Owned by: SMS-compliance docs, rival pricing pages

Search volumes: Outcraft SEO.xlsx, Focus Keywords tab (US). Journey stages and owners: the 31-query sweep in section 03.

03The 31 searches we tracked

Typed the way merchants type. Outcraft never showed up.

Each search was written from a merchant's point of view for Google or for an AI assistant, across discovery, consideration and decision. Click a row to see what it tells us. Direct competitors are highlighted in blue.

#
Stage
Search, exactly as typed
Vendors that surfaced
AI-calling vendor?
Outcraft?

Run 15 Oct 2026 through a web search engine (top 8–10 results), with Google US head terms cross-checked in Ahrefs (section 13). AI-assistant answers were not captured; Ahrefs Brand Radar is not on our plan.

04Who owns each stage

AI-calling vendors are invisible at discovery and dominant at decision.

Vendor appearances by category and stage

Share of all vendor mentions in that stage's searches

Discovery: 1 AI-calling mention in 47

Loyalty widgets, email recovery and subscription apps answer the merchant's symptom. Their answer is always more email, discounts or retries.

Consideration: voice appears, but via app listings

Shopify App Store listings (Hypothesis, Skiddly, Loop Voice, Callfy, Callsy) dominate “ai that calls abandoned carts shopify”.

Decision: AI calling takes 56% of mentions

Compare pages, pricing explainers and directory listings. Consio alone appears in 6 of 11 decision searches.

05Who we actually compete with

110 vendors surfaced. Five compete with us head-on.

Direct = sells AI that calls or messages ecommerce shoppers and appeared in at least two of the 31 searches. Pricing, ratings and funding as published by each company or its listing, 15 Oct 2026.

06Head-to-head scorecard

Outcraft is last or tied-last on 7 of 14 measures, and first on one.

Badges show Outcraft's rank among the six companies on each measure (1 = best). Blue column is Outcraft.

Competitive shape

Each axis scaled to the best company in the set. Tap names to compare.

Where we lead

The only company in the set publishing all four channels (voice, SMS, email and WhatsApp), and the second-largest integration catalog (21 pages, after Ringly.io's 41).

Where we trail most

Authority (DR 28, last), product pages for our own moments (0 of 4), decision-stage pages against this set (0), App Store presence (none) and case studies (tied last).

The closest rival

Consio: same ICP, all four moments, DR 60, App Store 5.0 (40 reviews), 21 customer stories and published pricing. It is the shape we need to beat.

07Authority and organic reach

Lowest authority of the six. Only Ringly.io turns authority into traffic.

Domain Rating and organic visits per month

Authority vs reach

Bubble size = ranking keywords; visits on a log scale

Consio: authority without traffic

DR 60 and 2,016 referring domains, but 9 ranking keywords and ≈12 visits a month. It wins shortlists without organic search; its authority is still unspent.

Our links are many but weak

1,167 referring domains yield DR 28; Callsy reaches DR 45 with 712. 64% of ours are dofollow, vs 74% for Ringly.io.

The gap to close

32 DR points to the competitor median of 60. Links, not more blog volume, are what hold back head terms.

Ahrefs batch analysis, all subdomains, all locations, 15 Oct 2026. Visits are Ahrefs estimates.

08Where the traffic comes from

84% of our organic visits come from buyers we no longer sell to.

Outcraft: 282 organic visits a month, by page

Ringly.io's six biggest pages: none about AI phones

Organic visits per month

India sends more than the US

India 106 visits, US 99, UK 32. Our ICP is US and Europe; the WhatsApp post skews to India.

ICP-aligned pages earn almost nothing

Our ecommerce retention post earns 14 visits; failed-payment and cart-recovery posts rank but earn ≈0.

Ringly.io buys attention first

Competitor pricing (Zendesk, Shopify Plus) and DTC brand pages attract Shopify operators, then sell them a phone agent. 91% of its traffic is US.

Ahrefs top pages, all locations, 14 Oct 2026. Off-ICP posts are on the Pruning List, so an early dip in total traffic is expected and intended.

09Content footprint

Ringly.io publishes 13× our pages, and 29% of them rank.

English pages in each sitemap, and pages that rank

Exact counts from XML sitemaps (competitors 15 Oct 2026; Outcraft 17 Sep 2026, before the /solutions consolidation). Translations listed separately. *Share of editorial, compare, use-case, case-study and programmatic URLs whose slug names an ecommerce topic; a slug-level proxy. Ranking pages: Ahrefs.

10How each rival wins

Four content strategies, and who they aim at.

Ringly.io
1,160
blog + programmatic pages

Volume and breadth

965 posts and 195 DTC-brand pages (“brands like AG1”). 51 comparison pages aimed at helpdesks and voice builders.

Consio
44
use-case, industry and audience pages

Product depth

16 use cases (30 pages with tag pages), 11 ecommerce industries, 21 customer stories and 104 help and update pages.

Callsy AI
809
translated pages, 6 languages

Decision stage and Europe

38 compare and alternatives pages including vs Outcraft, plus 41 glossary terms. Expanding into local trades.

Thoughtly
2%
of content about ecommerce

Horizontal platform

595 pages aimed at finance, insurance, mortgage and education. A sales-call rival, not a content rival.

Who compares against whom

Arrows = a compare or alternatives page naming that rival

Callsy is the only one targeting us

callsy-vs-outcraft exists in English and five translations. Nobody else in the set names Outcraft yet.

Ringly.io and Consio are the most-targeted

Ringly.io is named by Consio, Callsy and Loop Voice; Consio by Callsy and Loop Voice. That is where the buyer comparison happens.

Our 13 compare pages aim at the wrong rivals

Retell (3), Vapi (3), Synthflow (2), Bland (2): retired v3 rivals. The one Consio review is on the Pruning List.

11The positioning check

“Four moments, one sales team” is already on three competitor sites.

The Brand Book says no competitor covers failed payments and none runs all four moments. Their sitemaps say otherwise. Channel breadth is where we remain different.

Dedicated pages per moment

Channels on their product pages

Evidence

Consio: /use-case/failed-payment-save, /use-case/subscription-retention-and-cancellation-save, /use-case/post-purchase-upsell. Callsy: /use-cases/payment-reminders, /use-cases/subscription-renewal, /use-cases/win-back. Loop Voice: /use-cases/failed-payments, /use-cases/post-purchase-upsell. Outcraft: one combined /solutions page since 14 Oct; the four pillar pages (Topical Map rows 111–114) are not yet written. Channel coverage is taken from public product pages and needs product verification before sales relies on it.

12Google US reality check

On the head terms, none of the six ranks. Incumbents own them.

Google US results for the head terms behind our searches, including the sources cited in the AI Overview.

Ahrefs SERP overview, US, latest snapshot, 15 Oct 2026. DR in brackets. Long-tail natural phrasing has no stored SERP in Ahrefs.

13Price and speed claims

Merchants see $29–$49 a month everywhere. They see nothing from us.

Lowest paid entry price, USD per month

Ringly.io prices per call ($0.21–0.39 cited) and Consio adds $0.30 per AI minute; neither is a flat monthly fee.

Time-to-live claims

CompanyPublished claim
ConsioSame-day AI answering
FlipCXUnder an hour
Ringly.io5 minutes to a couple of hours
Callsy AIFree plan, self-serve install
Outcraft“Live in days”

Ours is the slowest stated claim. Brand Book §6 says to publish our real number if it is genuinely fast.

14How the web describes Outcraft

Every third-party description predates the ecommerce pivot.

Search engines and AI assistants build our entity from these sources. A competitor is currently writing one of them.

15Asset gap

Four gaps decide shortlists: moment pages, App Store, compare pages, price.

16What we should do, in order

Eight moves, ranked by shortlist impact per unit of effort.

Baseline to measure against (15 Oct 2026)

How to read progress

Re-run the same 31 searches monthly and count appearances by stage. Total organic visits will dip first, because 84% come from off-ICP posts on the Pruning List; that dip is intended. Success is appearing in decision-stage searches and on the head terms in section 12, not raw traffic.

17Method and limits

How every number here was produced.

Searches

31 natural ICP queries (10 discovery, 10 consideration, 11 decision; 21 written for Google, 9 as AI-assistant prompts, 1 for both), run through a web search engine on 15 Oct 2026, top 8–10 results logged. Every vendor surfaced was recorded and categorised (110 in total).

Authority and traffic

Ahrefs batch analysis, top pages, pages-by-traffic and SERP overview (US), all subdomains, 14–15 Oct 2026. Traffic figures are Ahrefs estimates; Outcraft's GSC remains the source of truth for our own numbers.

Content

Exact XML sitemaps: Consio 334, Ringly.io 1,395, Callsy 1,005, Thoughtly 595, Loop Voice 97 URLs (15 Oct); Outcraft 97 (17 Sep). Pages typed by URL path; translations separated; ecommerce share is a slug-level proxy.

Not covered

Answers inside ChatGPT, Perplexity, Gemini and Google AI Mode (Brand Radar not on our Ahrefs plan). Competitor claims, ratings, funding and prices are as published and not independently verified.