Forethought Pricing: The Real Cost Breakdown & Gotchas

Created time

Jul 11, 2026 11:48 AM

Title length (<60)

Author

Mike Heap

Last optimised

Ecomm?

How does Forethought pricing actually work?

⚡

TL;DR: Forethought charges a fixed platform access fee plus a usage component, but publishes neither figure. Every tier ends in a "Get a Quote" button with no self-serve price to check.

Forethought sells a platform as one bundle, with no menu to order line items from. In its own words, the pricing model is "a blend of platform access fees and an outcome-based pricing cost", with optional add-ons on top, a line that sits on the pricing page and is about as specific as the page gets. There is a fixed component (you pay to have the platform) and a usage component (you pay for what the AI does), and the split between them lands wherever your sales negotiation lands.

The catch is that neither component carries a number. Every tier (Team, Professional, Enterprise) ends in "Get a Quote", with no published platform fee, no published per-unit rate, and no self-serve signup that would let you find one.

The one usage rule the page does spell out is the overage clause: "additional usage charges may apply if your usage exceeds the plan's limits and purchase volume". But the limits themselves aren't published either, so you're agreeing to a ceiling you can't see.

The closest thing to real numbers comes from Vendr, which aggregates what its buyers actually paid. Vendr puts the median annual contract at $74,483 across 69 purchases, ranging from $35,670 to $151,400, with buyers negotiating an average of 22.4% off.

Forethought on the Vendr procurement marketplace

Vendr also publishes observed per-unit bands that never appear on Forethought's own page. Solve (the automated-resolution product) runs from $0.50 to $2.00 per conversation, Assist (the agent copilot) from $50 to $150 per agent per month, and Voice AI from $0.10 to $0.50 per minute or $2 to $8 per call.

One caveat, though. Vendr's page says its data was last updated February 2026, which is before Zendesk's acquisition of Forethought closed. So it's the best number available (take it with a grain of salt), but it predates the acquisition.

Here's the whole cost stack in one place. Where the real answer is "not published", that's what the cell says.

Component	Price	How it's billed	Notes
Platform access fee	Not published	Annual	Present on every tier. No figure disclosed
Usage ("outcome-based")	Not published	Vendor prices per resolution. Vendr's buyers pay $0.50 to $2.00 per conversation	The two do not describe the same unit
Assist (agent copilot)	$50 to $150/agent/mo (Vendr observed)	Per agent per month	Not shown on the pricing page
Voice AI	$0.10 to $0.50/min or $2 to $8/call (Vendr observed)	Per minute or per call	Not shown on the pricing page
Add-ons: Multibrand · Analytics API · Discover	Not published	Annual	Only three add-ons remain listed
Overages	Not published	On exceeding "plan's limits and purchase volume"	The limits themselves aren't published

A fixed platform fee has a real virtue: it's predictable once you've signed. But you have to sign to see it. Our own model at My AskAI is a flat per-ticket rate published on the page, so you can model your bill before you ever talk to us, the opposite starting position to a wall of "Get a Quote" buttons.

What is Forethought's "outcome", and why doesn't it match what buyers get billed for?

⚡

TL;DR: Forethought markets an outcome-based, per-resolution price, but the only observed buyer data, from Vendr, prices Solve at $0.50 to $2.00 per conversation. Those are different units: a conversation bills whether or not the issue resolves.

Nobody else reconciles this part, so I will. Forethought markets a resolution price. The pricing page calls it "an outcome-based pricing cost", and co-founder Deon Nicholas has said it directly on LinkedIn:

"we launched a new pricing model 6 months ago where we charge primarily for 'resolutions': where you pay if our AI agent can solve an issue for you, end-to-end, without needing a human touch."

He goes on: "the bulk of the value gets aligned to resolutions, or the 'outcome' you are paying for." He also confirms a "basic 'platform fee'" sits alongside it, and notably discloses no rate for either.

So the marketed unit is a resolution: you pay when the issue is solved. But the only observed buyer data, Vendr again, prices Solve at $0.50 to $2.00 per conversation, a different meter entirely.

Forethought markets a per-resolution price while the only observed buyer data, from Vendr, prices it per conversation at $0.50 to $2.00 each, a conversation billing whether or not the issue resolves.

A conversation and a resolution are not the same thing (that distinction runs through the whole invoice). A conversation bills whether or not it ends in a solved issue. A resolution, by the marketed definition, only counts when the AI closes the loop with no human.

Depending on how many of your conversations actually resolve, those two denominators can produce very different invoices for the same volume of traffic, and I've seen that gap catch buyers out. And you cannot tell which one you'll be billed on without going through sales, because the page defines neither the outcome, nor the rate, nor how a resolution is verified.

It's a sharper gap when you notice that Forethought's new parent defines its unit and Forethought doesn't. Zendesk, which now owns Forethought, publishes an actual definition of an autoresolve, which fires after a 72-hour quiet period, meaning the ticket stays closed and the customer doesn't reopen the conversation.

That's a testable, countable event. I'd want that kind of clarity before signing anything. Forethought's page names the unit and defines nothing.

Two things I'm deliberately not going to claim, because the internet keeps getting them wrong. First, this isn't a post-acquisition repricing. Nicholas announced the outcome-based model well over a year before Zendesk's deal, so when Zendesk uses the same "outcomes" language now, the two have simply landed on it independently.

Second, an old figure floating around the web has Forethought at $0.12 per deflection. That's a deflection denominator, and Vendr's $0.50 to $2.00 is a conversation denominator. They measure different things, and anyone presenting the pair as a price rise is comparing two different meters.

The resolution unit makes me wary for a reason that comes straight from running My AskAI. Most of what makes a resolution rate climb is your own work: updating knowledge, wiring up the API and tool connections, tuning guidance, setting up triage, and running the weekly QA loop that catches what the AI got wrong.

An outcome-priced vendor bills you more as that number goes up, which means you're paying them for improvements your own team produced. We deliberately went the other way with a usage-based, flat rate per ticket, so as the AI gets better the upside stays with you and doesn't flow back to us as a higher bill.

What are the real Forethought tiers, and what's gated behind each?

⚡

TL;DR: The current tiers are Team, Professional and Enterprise. The old "Basic" ladder that most pages still print is gone, and multilingual, QA rubrics, brands and triage models now start at Team, so the entry tier includes more than the stale captures show.

The tiers are Team, Professional and Enterprise, and the old "Basic" name is gone. Capterra still files the whole product under its pre-2021 name, "Agatha", and most third-party pages still print the old "Basic" ladder. The bigger change is what's underneath, where a lot of what used to be paywalled has moved down into the entry tier.

Here is what sits in each tier on the live page today:

Team: Chat and Mobile channels. Autoflows, custom actions, CSAT collection. AI QA with 1 rubric included. 2 brands included. Sentiment, intent and spam triage models (3 included). Multilingual support. Integrations, an insight dashboard, permissions/security/logs, and standard support.

Professional: everything in Team, plus Email, Voice and Slack channels. 5 QA rubrics, 20 brands, and 6 triage models (with the ability to build custom handoff models). AI-surfaced insights and "Ask the expert" support.

Enterprise: everything in Professional, plus the Solve API, 20 QA rubrics, knowledge-gap detection with AI article creation, autoflow gap detection and generation, the Analytics API, enterprise security and governance controls, and advanced support.

Every stale page I've seen gets the gating shifts backwards. Multilingual used to be a Professional-and-up feature. It's now included at Team.

The entry tier is no longer chat-only. It has mobile too. QA rubrics, brands and triage models that used to be add-ons or higher-tier perks now start at Team.

And the add-on list has shrunk to just three: Multibrand (up to 20 brands, each with its own AI personality), the Analytics API (conversation-level data across chat, email, voice, API and Slack, plus AI QA scores), and Discover (knowledge-gap detection from real ticket data, AI-recommended article creation, and suggested automation).

None of this comes with a price, so "what's included at Team" tells you what you're negotiating for without telling you what you'll pay for it. But if what you know about Forethought's tiers came from an older write-up or an AI summary, start here, because the ladder has changed underneath it.

What does Forethought cost at typical volumes?

⚡

TL;DR: At Vendr's $74,483 median (about $6,207 a month), the effective cost lands near $0.89 per resolved ticket at 10,000 tickets a month and around $0.18 to $0.25 at 50,000. Below the 20,000-ticket floor nobody actually buys.

Let me put real numbers against real volumes. Forethought is fixed-annual, so the volume you run doesn't create per-resolution variability the way it does with a usage-metered vendor. It only changes your effective cost per resolved ticket.

And the figure I'm modeling is Vendr's median: $74,483 a year, which works out to about $6,207 a month. Forethought publishes no list price, so that figure is a median of what real buyers negotiated, which I read throughout as "what the middle buyer paid".

For the resolution rate, I'm anchoring the headline to the field median from our own benchmark of 195 deployments across roughly 55 vendors, which lands at about 70% on a resolution metric (mean 66.6%, P25 56%, P75 80%). I show a 50% column alongside it as a deliberately pessimistic floor, well under both the benchmark and what a competent deployment should expect.

Volume	Annual (median contract)	~Monthly	$/resolved @70%	$/resolved @50%
1,000 tickets/mo	~$74,483	~$6,207	~$8.87	~$12.41. Below the product's 20,000-ticket floor, so not a real option
10,000 tickets/mo	~$74,483	~$6,207	~$0.89	~$1.24
50,000 tickets/mo	~$74,483	~$6,207	~$0.18	~$0.25

At 1,000 tickets a month

At 1,000 tickets a month you're well below the floor (the product is built for 20,000+ historical tickets and 2,000+ a month), so nobody actually buys here.

If someone did, a $74,483 contract against ~700 resolved tickets a month works out to nearly $9 a resolved ticket. That tells you exactly why the floor exists. The fixed platform fee only makes sense once there's enough volume underneath it to divide by.

At 10,000 tickets a month

This is the volume where I'd say Forethought becomes a plausible buy, and where the median contract divides into something reasonable. That's about $0.89 per resolved ticket at a 70% resolution rate, and about $1.24 at the conservative 50% floor.

Because the bill is fixed, those are the numbers whether you have a light month or a heavy one, and you've pre-paid the year. The variability that a per-resolution vendor pushes onto your monthly invoice, Forethought pushes onto your annual negotiation instead.

At 50,000 tickets a month

At high volume the fixed contract works in your favor, because the same ~$6,207 a month spread across far more resolved tickets drops the effective rate to roughly $0.18 to $0.25 per resolved ticket. This is the end of Forethought's model I'd point a high-volume team to.

Because the bill doesn't move, your cost per resolution falls as volume and resolution rate rise. That's the same direction a flat per-ticket model like ours runs, and the opposite of a per-resolution vendor whose bill climbs as it succeeds.

As the AI resolution rate climbs from 50% to 90%, Forethought's fixed-annual cost per resolved ticket falls from about $1.24 to $0.69, while a per-resolution meter stays flat at about $1 per outcome.

Forethought's "success tax" is real, and it lands at the contract level: the 20,000-ticket floor and the annual commitment you make before you know it works on your tickets. What you won't find is a meter that charges you more with every resolution.

What does the Forethought pricing page not tell you?

⚡

TL;DR: Three things sit off the page: a 20,000-ticket eligibility gate, annual-only terms with renewal uplifts you have to negotiate out, and a 30 to 90 day go-live with no way to test on your own tickets first.

Three things sit off the page that change the deal, and none of them are dollar figures. They're the hidden costs of AI customer service I watch surface halfway through a rollout.

Three things Forethought's pricing page leaves off: a 20,000-ticket eligibility gate, annual-only terms with renewal uplifts, and a 30 to 90 day go-live with no way to test on your own tickets first.

Hidden cost 1: the 20,000-ticket eligibility gate

Forethought works best with 20,000+ historical tickets and asks for 2,000+ email or chat tickets a month on an ongoing basis. That's an eligibility gate, and for a lot of readers it's the line that decides whether Forethought is a real option at all.

If your team isn't doing that volume, the "Get a Quote" button is decorative, and you can't really buy this product at any number. That's not a knock on Forethought (it's built for scaled CX orgs), but a pricing page that lets a 5,000-ticket-a-month team fill in a form is wasting their afternoon.

Hidden cost 2: annual-only terms and renewal uplifts

Payment is annual, with no monthly option to hedge with. And Vendr's buyer notes flag that price uplifts at renewal and auto-renewal clauses are things buyers routinely negotiate out of the standard contract, which is another way of saying they're in it by default.

If you don't have someone reading the renewal terms with a red pen (and I'd make sure you do), the second year can cost more than the first for reasons that have nothing to do with your usage.

Hidden cost 3: 30 to 90 days to go live, and no way to test first

Most teams are live in 30 to 90 days, and reviewers consistently flag the absence of a simulation or testing mode. I'd want one, personally. There's no sanctioned way to run Forethought's AI against your historical tickets and see how it would have done before you commit.

G2's own pricing insights put time-to-implement at about 2 months and return-on-investment at about 11 months. Stack that up and the commitment is clear: you sign an annual contract, wait a couple of months to go live, and wait the better part of a year to see the payback, all before you have real evidence it works on your tickets.

That's the trade you're making for a fixed, predictable line item.

What do real Forethought customers say about the invoice?

⚡

TL;DR: Public billing complaints are almost nonexistent: of roughly 165 G2 reviews about four mention cost, and Reddit is effectively silent. That silence is what a sales-gated, NDA-shaped, annual-contract motion produces, and it says nothing about whether the price is fair.

Almost nothing, and the almost-nothing is itself the finding. I went looking for billing complaints the way I would for any pricing post: Reddit, G2, Capterra, Hacker News.

Of roughly 165 G2 reviews, about four say anything about cost. Capterra's eleven don't mention price once, and the word-on-the-street you'd expect on Reddit and Hacker News is effectively silent.

The two G2 reviews that do touch cost both name the actual pattern:

"Forethought AI delivers solid automation, but its pricing is a major downside. The pricing can make budgeting difficult…" — William G., Small-Business, G2

"How split up the features are, alltogether they provide a great service but it adds up." — Vincent C., Mid-Market, G2

Vincent's point is the more useful one: with the platform split across Solve, Triage, Assist, Discover and QA, the invoice is the sum of several parts, and "it adds up" is exactly what you'd expect from a suite priced by the module. William's is the budgeting problem restated. When the rate is set by negotiation and never published, I'd budget next year's spend wide, because it's hard to forecast.

There's a real tension in the discount data, too. G2's on-page pricing insights report an average discount of about 11%, while Vendr's 69 buyers report average savings of 22.4% (with individuals landing anywhere from 19% to 57%).

Both can be true at once, and the spread is the story: when the list price is a secret, what you pay reflects how hard you negotiate more than what the product costs to run. Two companies buying the same thing in the same quarter can be 30 points apart on price.

Reported Forethought discounts spread widely: G2 buyers average about 11% off, Vendr's buyers average 22.4%, and individual buyers have reached 57% off.

The plain reading of near-zero public billing discussion for a company that's raised over $115M is straightforward: this is what a sales-gated, NDA-shaped, annual-contract motion produces. An absence of complaints here says nothing about whether the price is fair. It only tells you that nobody who bought it can compare invoices with anybody else.

Does Forethought have a free trial, and what are the contract terms?

⚡

TL;DR: There's no free trial, only a sales-led Proof of Value pilot run on your data, and contracts are annual only. Discounts are large and reliable: Vendr's buyers report 15% to 30% off initial quotes, and individuals have reached 19% to 57%.

No free trial. In Forethought's own words: "Yes, we offer a Proof of Value (POV) instead of a traditional free trial. We work with you and your data to demonstrate how swiftly and effortlessly Forethought can be deployed and utilized to showcase the platform's effectiveness in real-world scenarios."

In plain English, that's a sales-led pilot: you work through it with their team on your data, with no self-serve account to spin up on a Tuesday and kick the tires. There's no free plan on any tier and no sandbox, and though some legacy "Solve Lite" and "Triage Quickstart" trial pages still exist, they aren't the current motion.

Payment is annual only, and the discounts are large and reliable. Vendr's buyers report the levers clearly: 15% to 30% off initial quotes when you bring competitive context, 10% to 20% for committing to 2 to 3 year terms, 10% to 15% better per-product pricing when you bundle, and 20% to 35% lower when you have multiple alternatives in play.

Timing is a lever too: engage 60 to 90 days before go-live, start renewal conversations 90 to 120 days before expiry, and use quarter- and year-end (March, June, September, December) for leverage. Individual buyers have secured 19% to 57% off, and one landed 50% in year one and 32% in year two.

A discount that large and that dependable tells you something about the list price. A number you can reliably knock a third off was only ever an opening position.

That's the buyer's real disadvantage with opaque, sales-only pricing: you're negotiating against a figure you can't see, and the vendor knows what it'll settle for while you're guessing.

For contrast, this is the exact objection our model is built to remove. At My AskAI the pricing is published on the page, and the trial is a proper one: 30 days, every feature unlocked, unlimited tickets, no credit card.

You prove it on your own tickets before you pay anyone, and you don't pre-commit to a year on the strength of a sales-led pilot. That difference is most of why price-sensitive teams end up on our page in the first place.

How does Forethought pricing compare to alternatives at the same volume?

⚡

TL;DR: At 10,000 tickets a month, Forethought's median works out to about $6,207 a month, roughly level with Intercom Fin, cheaper than Zendesk's own AI, and about 4.8× My AskAI's ~$1,299. Its edge is predictability: the bill doesn't climb as the AI resolves more.

Line Forethought up against the field at 10,000 tickets a month, using 50% resolution as the common comparison line. One caveat up front, the same one this whole post is about: Forethought's cell is a median contract value, while every other cell is computed from a published rate.

That's not an apples-to-apples row, and pretending it is would be the exact error the rest of these pricing pages make. So read Forethought's number as what the middle buyer paid, a negotiated median that no published rate stands behind.

Vendor	Pricing model	Cost at 10k (50%)	$/resolved (50%)	$/resolved (70%)
Forethought	Fixed annual, sales-only (median contract)	~$6,207	~$1.24	~$0.89
Zendesk AI agents	Suite seats + $1.50/autoresolve	~$10,500	~$2.10	rises with success
Intercom Fin	$85/seat + $0.99/outcome	~$6,650	~$1.33	rises with success
eesel AI	PAYG $0.40/task (1 ticket = 1 task)	~$3,000 annual	~$0.80	~$0.57
My AskAI	Flat $0.10/ticket, no seat fees	~$1,299 (Scale)	~$0.26	~$0.19

A few reads on that table. Forethought's median lands between Intercom Fin and Zendesk's own AI at this volume: cheaper than Zendesk AI agents, roughly level with Fin, and about 4.8× My AskAI ($6,207 against $1,299). (Intercom has rebranded its AI to Fin, and Salesforce has signed to acquire it, though that hasn't closed, so keep Fin and Salesforce's Agentforce as separate things in your head.)

In Forethought's favor, the fixed-annual model wins on one axis the per-resolution vendors lose on: predictability. Forethought's CFO gets a single number for the year, and it does not climb as the AI gets better.

The five tools compared group into three pricing models: Forethought's fixed annual contract, per-outcome-plus-seat meters that climb with success (Zendesk AI, Intercom Fin), and flat per-ticket rates you can forecast (My AskAI, eesel AI).

Zendesk's and Fin's $/resolved both rise with success, because you pay more each time the AI resolves more, so the better it works, the more it costs. Forethought doesn't do that.

I'm not going to claim we're always cheaper. At the low end of Forethought's negotiated band, a very large, well-run deployment can reach a competitive cost per resolved ticket. The counter is narrower: our unit is one you can forecast before you sign, because it's a flat rate per ticket with no seat fees, and the price is on the page.

Video preview — I Let AI Agents Resolve 10,000 Support Tickets, Here's How Much It Cost

You can model your My AskAI bill in a spreadsheet tonight. You cannot model your Forethought bill without a sales call.

Is Forethought actually worth the money?

⚡

TL;DR: Forethought is worth it if you clear the 20,000-ticket floor, want the full Solve, Triage, Assist, Discover and QA suite, and value one fixed annual line over a usage curve that climbs. It's the wrong buy if you're under the floor, need to be live this quarter, or want to test on your own tickets before committing a year.

The price only makes sense against the results, so start with what you're actually buying. Forethought's homepage claims deflection of up to 98%, but its own published case studies land more realistically in the 44% to 87% range, where Grammarly reports 87% and YAZIO around 80%. Where you land in that range decides the math above: your effective cost per resolved ticket depends entirely on which end you hit, and the contract is fixed regardless of where you do.

Zendesk completed its acquisition of Forethought on 26 March 2026 (announced 11 March), in what it called its largest deal in nearly 20 years, confirmed in the completion release and covered at announcement. This is where speculation runs wild, so here's what the paperwork actually says.

The completion release says nothing about Forethought's pricing or packaging changing, and nothing about its standalone availability to non-Zendesk customers. The "remains available to non-Zendesk customers" line came from the March 11 announcement only and hasn't been restated since. And any "expect a 15% to 25% uplift post-acquisition" claim you'll find traces back to competitor blogs, with no customer or vendor statement behind it, so I'm not repeating it as fact.

So the real answer to "did pricing change after Zendesk bought them?" is: no announced change, and unknown in practice.

So who is it actually for?

✅

Forethought is worth it if:

You have 20,000+ historical tickets and 2,000+ a month, so you clear the floor.

You're already committed to Zendesk, which is now the parent company.

You need the full five-agent suite of Solve, Triage, Assist, Discover and QA, where deflection is only one piece.

You'd rather have one fixed annual line than a usage curve that climbs as your AI improves.

You have the procurement muscle to take 20% to 35% off list, because the list price is a starting position.

❌

Forethought isn't worth it if:

You're under the ticket floor, so most smaller teams simply aren't the buyer here.

You need to be live this quarter, and 30 to 90 days is too long to wait.

You want to test the AI on your own tickets before committing to a year (for a lot of teams, this is the dealbreaker).

You want to know the price before you take a sales call.

You'd rather not tie your AI layer to Zendesk's platform direction.

If you're doing this much homework to find out whether you can afford Forethought, the answer is usually that it's a stretch, and the teams it fits rarely have to wonder whether they qualify. If that's you, the smart move is to look at a model you can price yourself.

My AskAI runs inside your existing helpdesk on a published $0.10-a-ticket rate, is SOC 2 Type II certified and GDPR-compliant, and gives you 30 days with everything unlocked and no card to find out whether it works on your tickets, before you commit a budget line to anyone, us included.

For the full feature picture, the Forethought complete guide covers what the platform actually does, and the Forethought alternatives roundup scores the substitutes on transparent pricing and setup time.

And if you want to see what the flat-rate model looks like on a real invoice: Honeygain holds a 90% resolution rate across roughly 3,400 Zendesk tickets a month, and Apartment List resolves 76% of theirs at 97% AI CSAT. There are more in our case studies.

FAQs

How much does Forethought cost?

There's no published price. Forethought lists three tiers, all ending in "Get a Quote", and no dollar figure appears on any of them. The best available number is Vendr's median annual contract of $74,483 across 69 buyers (range $35,670 to $151,400), stamped February 2026, before the Zendesk acquisition closed, so treat it as a pre-close figure that may not hold today.

What counts as an "outcome" in Forethought's pricing?

Forethought doesn't define it. The pricing page calls the model "outcome-based" and co-founder Deon Nicholas describes charging "primarily for 'resolutions'", but the page carries no definition of what qualifies as a resolution, no rate, and no verification method. Its parent company Zendesk does define its own unit (a 72-hour quiet period before an autoresolve counts), while Forethought names the unit and defines nothing.

Does Forethought charge per conversation or per resolution?

Forethought markets per resolution, but the only observed buyer data, from Vendr, prices Solve at $0.50 to $2.00 per conversation.

Those are different denominators: a conversation bills whether or not it resolves. Which one lands on your invoice isn't something you can determine without a sales call, because the page states neither the unit definition nor the rate.

Does Forethought publish its pricing?

No. Three tiers, three "Get a Quote" buttons, zero prices. The pricing page ranks first for "forethought pricing" and contains no numbers at all, which tells you most of what you need to know about how Forethought sells.

Is there a minimum ticket volume to use Forethought?

Effectively yes. Forethought is built for 20,000+ historical tickets and expects 2,000+ email or chat tickets a month on an ongoing basis. That volume works as a real eligibility gate: teams below it aren't the intended buyer, whatever the "Get a Quote" button implies.

Does Forethought have a free trial?

No traditional trial. In its own words, Forethought offers "a Proof of Value (POV) instead of a traditional free trial", which in practice is a sales-led pilot run with your data, with no self-serve account to spin up. There's no free plan and no sandbox. (For comparison, My AskAI's trial is 30 days with all features unlocked, unlimited tickets and no credit card.)

Can I cancel mid-contract, and are Forethought contracts annual?

Contracts are annual only, with no monthly option. Standard terms carry renewal uplifts and auto-renewal clauses that buyers routinely negotiate out, which means they're in the default contract unless you remove them. Read the renewal language before you sign (I would, every time), because the second year can cost more than the first for reasons unrelated to your usage.

What discounts can I negotiate on Forethought?

Large ones, reliably. Vendr's buyers report 15% to 30% off initial quotes with competitive context, 10% to 20% for 2 to 3 year commitments, and 20% to 35% lower when multiple alternatives are in play.

G2's insights show an average discount around 11%, while Vendr's sample averages 22.4%, with individuals reaching 19% to 57%. The size and consistency of those discounts tell me the list price is only ever an opening position.

Did Forethought's pricing change after Zendesk acquired it?

No announced change. Zendesk completed the acquisition on 26 March 2026, and its completion release is silent on Forethought's pricing and packaging.

The "outcome-based" wording predates the deal by well over a year, so it isn't a post-acquisition repackage. Any specific "expect an X% uplift" prediction comes from competitor commentary with no Forethought or Zendesk statement behind it, so the accurate answer is that post-close repricing is unknown.

How does Forethought pricing compare to Intercom Fin and Zendesk AI?

At 10,000 tickets a month and a 50% resolution rate, Forethought's median works out to roughly $6,207 a month, versus about $6,650 for Intercom Fin and about $10,500 for Zendesk's own AI agents. Forethought lands between the two, cheaper than Zendesk AI and roughly level with Fin, and about 4.8× our own ~$1,299 at the same volume.

Which AI support tool gives the best value per conversation?

It depends on whether you can forecast the unit you're billed on. The best value goes to a model where you can predict your own usage and see the rate before you sign, a flat per-ticket or per-conversation rate you can model in a spreadsheet.

Forethought is the hardest case here: it markets a resolution meter, Vendr observes a conversation meter, and neither is priced, so a real "value per conversation" can't be computed without a sales call. A published, flat per-ticket rate (My AskAI is $0.10) is the one you can actually check yourself.

Are there extra charges for voice or multilingual support?

Multilingual support is included at the Team (entry) tier, which corrects most pages still claiming it's gated higher up. Voice is a Professional-tier channel, and Vendr observes Voice AI at $0.10 to $0.50 per minute or $2 to $8 per call, though, like everything else, no rate appears on Forethought's own page.

Forethought Pricing: The Real Cost Breakdown Behind the Quote

How does Forethought pricing actually work?

What is Forethought's "outcome", and why doesn't it match what buyers get billed for?

What are the real Forethought tiers, and what's gated behind each?