Ada pricing

About

Ada is an AI agent platform for customer service that automates support across chat, email, voice, and SMS. The product is positioned around resolving customer inquiries end-to-end with AI agents rather than routing them to human staff, and the company markets ROI outcomes such as reduced average handle time and human labor hours saved per month. This places Ada squarely in the emerging market for agentic AI workflows, where the cost driver is resolved interactions rather than seats.

Ada is a Toronto-based company led by co-founder and CEO Mike Murchison. It reached Canadian unicorn status with a $130M Series C at a $1.2B valuation in May 2021 (led by Spark Capital, with Tiger Global, Bessemer, Accel, and FirstMark participating), bringing total funding to roughly $200M. The company anchors its enterprise positioning with named customers such as Pinterest, Monday.com, Cebu Pacific, and IPSY, alongside GDPR, HIPAA, and AICPA SOC compliance.

Ada targets high-volume support organizations. Its pricing page self-qualifies prospects by stating the company is a great fit for organizations with at least 300,000 annual customer service conversations, and the lead-capture form offers contact-volume brackets that extend past 100 million conversations per year — an unusually direct enterprise gate.

Ada publishes no public price table, so this analysis combines Wayback-verified packaging history with third-party purchase data (Vendr) and teardowns (eesel, fin.ai) to recover indicative pricing. Those figures — roughly $1.00-$3.50 per resolution and a ~$70K/yr median contract — are clearly marked as third-party-indicative throughout and are not official Ada rates.

Pricing summary : outcome-based AI customer service pricing, quoted by sales

Ada uses an outcome-based, resolution-priced, sales-quoted model. There is no public price table, no advertised free tier, and no self-serve trial. The /pricing page is a book-a-consultation form, and a quote is produced only after a scoping conversation with sales. Pricing scales with resolved-conversation volume plus the feature tier, not with seats.

What the public page shows, plus what third-party data fills in:

• No published rates on the page. Every dollar figure is sales-gated. Third-party purchase data (Vendr) and teardowns (eesel, fin.ai) indicate roughly $1.00-$3.50 per resolution, a ~$70K-$72K median annual contract (range ~$30K to $250K+), and per-conversation overages around $0.10-$0.50 above contracted volume.
• Volume-scoped qualification. Ada states it is a great fit for companies with at least 300,000 annual customer service conversations, and the form collects expected annual contact volume in brackets up to more than 100 million.
• Outcome framing. Marketing emphasizes resolution and efficiency outcomes (reduced handle time, labor hours saved) rather than seats, consistent with an outcome-based pricing approach metered on resolutions. This is the same value metric Intercom’s Fin popularized — see our Intercom Fin pricing teardown.

What makes this different: Ada commits hard to the resolution as the value metric and refuses to publish a list price, turning the pricing surface into a sales-led qualification funnel that discloses ROI proof points and a minimum-scale fit bar but not a single rate.

Pricing by product

Ada publishes no plan grid or per-unit rates on its page. The tables below combine what is publicly visible with third-party-indicative figures (Vendr purchase data, eesel and fin.ai teardowns, accessed 2026-06-07). Indicative figures are flagged explicitly and are not official Ada rates.

Ada AI Agent platform

Indicative rates (third-party, not vendor-published)

A material caveat on the metered unit: Ada markets the charge as outcome-based (“only pay for conversations that were actually resolved”), but third-party teardowns (eesel, fin.ai) argue Ada effectively bills per conversation — charged whether or not the AI resolves the issue. Buyers should confirm in writing whether their meter is per resolution or per conversation, because the two diverge sharply at low automation rates. See our explainer on usage-based pricing models for why the value-metric definition matters.

Sales motions across products: sales-led for all access — there is no self-serve or PLG path, and no published list price.

Hidden costs : what a sales-gated quote leaves undisclosed

Ada discloses no rates on its page, so any bill estimate must lean on third-party-indicative figures. The archetype below uses the recovered per-resolution band (~$1.00-$3.50) and is illustrative, not a quote — the real number comes only from sales.

A mid-market support team handling 30,000 resolved conversations per month (360,000/year, just over Ada’s fit bar):

Annualized, that lands near $810K at the midpoint rate — well above the ~$70K Vendr median, which illustrates how heavily the per-resolution rate and the resolution-vs-conversation definition swing the bill. A buyer at the $1.00 end of the band with negotiated volume pricing would pay a fraction of that. The unknowns that still move the total materially:

• Whether the meter is per resolution or per conversation (the latter charges for failed/escalated chats too).
• Voice automated resolutions and SMS carry separate charges on top of the base meter.
• Onboarding, implementation, and professional-services fees are undisclosed and typically front-loaded in year one.

Want to estimate your own Ada bill? A dedicated Ada pricing calculator is not yet published. In the meantime, our resolution-based billing guide walks through modeling a per-resolution meter, and the pricing blueprint lists comparable AI customer-service vendors that do publish rates.

Pricing evolution : from three named tiers to outcome-based resolution pricing

Cadence

Tracked range: 2022-07 to 2026-06 (Wayback for 2022-07 to 2023-09; live capture 2026-06). 2024-2025 snapshots were not re-captured this pass; quarters not listed showed no verified packaging change.

Notable changes

• 2022-07 to 2023-08 — Three-tier Core / Advanced / Pro plan grid, all “GET A QUOTE,” with add-ons (Social Channels, Additional Bots, Data Export API). Source: Wayback web.archive.org/web/20220702121947/https://www.ada.cx/pricing.
• 2023-09 — Pivot to outcome-based, resolution-priced model and the Generative/Scripted split. Source: Wayback web.archive.org/web/20230924145920/https://www.ada.cx/pricing.
• 2026-06-07 — Pricing page reduced to a sales-qualification form with a 300,000-conversation fit bar and ROI stats; no tiers or rates shown.

The 2023 pivot to outcome pricing in detail

The shift Ada made between August and September 2023 is the page’s defining inflection. The 2022-era page sold capability tiers (Core “Answer,” Advanced “Action,” Pro “Anticipate”) — a classic feature-laddered SaaS structure where the buyer picks a tier and gets a quote. The September 2023 page replaces that entirely with a single value-metric promise: “Get access to all the capabilities you need while only paying for the conversations that were actually resolved.” This put Ada in the same outcome-based camp as Intercom’s Fin, which had launched per-resolution pricing earlier that year. The move also consolidated the product around two builder modes — Generative (knowledge-base grounded, no training) and Scripted (NLU drag-and-drop) — signaling Ada’s bet that generative AI, not hand-built flows, was now the default. The trade-off is transparency: the tiered page at least named what each step included, while the outcome page discloses neither the resolution rate nor any minimum, pushing all price discovery into the sales call.

What’s unique : pricing page as a scale-qualification funnel

1. The pricing page sells qualification, not prices. Ada’s /pricing URL contains no rates at all — it is a consultation form with ROI proof points and a contact-volume selector, turning the pricing surface into a lead-scoring funnel. This is a sharper version of the sales-led motion than peers who at least publish a “starting at” anchor.

2. An explicit minimum-scale fit bar. Stating “at least 300,000 annual customer service conversations” is an unusually direct way to disqualify small buyers before any sales contact — a pricing lever applied before a single dollar is named.

3. Resolution as the committed value metric. Ada threw out its Core/Advanced/Pro feature tiers in 2023 to bill on resolved conversations, aligning charges with the outcome the buyer cares about. It is one of the clearest commitments to per-resolution pricing in the customer-service category, alongside Intercom’s Fin.

4. A resolution-vs-conversation ambiguity worth scrutinizing. Ada markets “pay only for resolved conversations,” yet third-party teardowns argue the practical meter is per conversation. That gap between the marketed and the billed unit is itself distinctive — and a reason buyers should pin the definition down in the contract, as we explain in our outcome-based pricing analysis.

Strengths & weaknesses

Billing UX : the only public controls are lead-capture fields

The pricing page exposes no billing dashboard or plan picker — only a sales-qualification form. The visible named controls are:

• Business Email field — required lead-capture input.
• Company field — required lead-capture input.
• Expected annual contact-volume selector — required dropdown with brackets from 0–99,999 up to more than 100 million conversations (plus an “I don’t know” option).
• “Speak to an expert” CTA — the sole action; submits the form to route the prospect to sales.

No invoicing, usage-meter, budget-cap, or self-serve billing control is publicly visible.

Strategic wins : qualifying buyers before quoting

1. Using the pricing page to disqualify sub-scale buyers

By stating a 300,000-conversation fit bar up front, Ada filters out small prospects before sales engages, concentrating effort on high-value accounts. This mirrors the gated, enterprise-first motion seen across sales-led AI vendors.

2. Leading with outcomes, not features

Framing the page around ROI metrics rather than feature checklists aligns the buying conversation with value, which supports premium outcome-based pricing once a quote is reached.

3. Front-loading trust signals on the pricing surface

Placing GDPR, HIPAA, and AICPA SOC badges directly on the consultation page de-risks the enterprise buying decision before sales engages, a tactic that pairs well with the agentic AI pricing motion Ada runs.

Areas to improve : zero transparency raises buyer friction

1. Publish a starting per-resolution band

Third-party sites already quote ~$1-$3.50 per resolution and a ~$70K median — so the figures are public, just not from Ada. Naming an official starting band would let Ada control its own pricing narrative and capture the research-stage traffic that now flows to aggregators and competitors instead.

2. Resolve the resolution-vs-conversation ambiguity publicly

The single biggest buyer anxiety is whether failed/escalated chats are billed. Defining “resolution” precisely on the page — as the resolution-based billing guide recommends — would differentiate Ada from peers whose definitions are murky and reduce contract-stage friction.

3. Offer a self-serve estimate path

A volume-to-estimate calculator (even gated) would let prospects self-qualify on price as well as scale, shortening the sales cycle. Pairing it with the published per-resolution band would let buyers model spend before the first call instead of discovering the rate only in the redline.

Key takeaways

1. Repackaging on the value metric is a one-way door worth taking. Ada replaced three capability tiers with a single resolution meter in 2023, betting the outcome is what buyers will pay against — a clearer alignment than feature ladders.
2. Scale qualifiers are a pricing lever. Publishing a 300,000-conversation fit bar pre-segments demand and protects sales capacity before any rate is named.
3. Gating the rate does not gate the data. Vendr, eesel, and fin.ai already publish Ada’s ~$1-$3.50/resolution and ~$70K median — so opacity just cedes the pricing narrative to third parties.
4. Define the value metric precisely or buyers will distrust it. The resolution-vs-conversation ambiguity is Ada’s biggest avoidable friction point; a published definition would convert skeptics.
5. Indicative pricing can be recovered honestly. Triangulating aggregator purchase data with Wayback packaging history reconstructs a credible price band without inventing a single figure.

UBP implications

1. Outcome-based models often hide behind sales gates. Ada meters on resolutions yet publishes no rate, showing how outcome pricing and price opacity frequently travel together and limit market price discovery.
2. Volume qualification is a precursor to usage pricing. Bracketed contact-volume selectors are the same dimension a usage quote is built on, surfaced early as a qualification tool — a pattern usage-based vendors can copy to pre-size deals.
3. The value-metric definition is the real contract. For resolution pricing, whether failed conversations count is worth more than the headline rate; UBP teams should specify the metric as tightly as the price.

Sources

• Ada pricing page (accessed 2026-06-07)
• Ada website (accessed 2026-06-07)
• Ada blog (accessed 2026-06-07)
• Ada Series C announcement — $130M at $1.2B valuation (accessed 2026-06-07)

Third-party indicative pricing in this analysis is drawn from Vendr, eesel, and fin.ai (cited inline) and recorded with provenance in the research record; per the Sources contract, aggregator and teardown sources are not listed here as primary sources.

Bottom line

Ada is an enterprise AI customer-service platform that bills on resolved conversations but publishes no rate — its /pricing page is a consultation funnel with a 300,000-conversation fit bar and ROI stats. Triangulating Wayback packaging history with third-party purchase data puts Ada near $1-$3.50 per resolution and a ~$70K median annual contract, all marked third-party-indicative rather than vendor-confirmed. The 2023 pivot from Core/Advanced/Pro tiers to outcome pricing is the story; the unresolved resolution-vs-conversation definition is the risk.

Want to compare Ada against other AI customer-service pricing? Browse the pricing blueprint.

Pricing timeline : Major events on a vertical axis

Each milestone below corresponds to a public pricing change, product launch, or material adjustment. Major events use a filled marker; minor adjustments use a faded one.

Pricing page is a sales-gated consultation form

Jun 2026

Ada's /pricing surface now shows no price table or tiers — only a book-a-consultation lead form, a 300,000 annual-conversation fit qualifier, and ROI stats. All dollar figures are sales-quoted; third-party aggregators indicate roughly $1-$3.50 per resolution and a ~$70K/yr median contract.

Pivot to outcome-based, resolution pricing

Sep 2023

Between the 2023-08-16 and 2023-09-24 snapshots Ada replaced the Core/Advanced/Pro tiers with an outcome-based model headlined 'pay only for the conversations that were actually resolved,' and re-split the product into Generative and Scripted AI builders. Source: web.archive.org/web/20230924145920/https://www.ada.cx/pricing

Three named tiers: Core / Advanced / Pro

Jul 2022

Wayback snapshots from 2022-07 through 2023-08 show a three-tier plan grid — Core (Answer), Advanced (Action), Pro (Anticipate) — each carrying a GET A QUOTE button with no public dollar amount. The page also listed add-ons (Social Channels, Additional Bots, Data Export API). Source: web.archive.org/web/20220702121947/https://www.ada.cx/pricing