Once you strip bundled Now Assist and Impact support out of the new tier quote, most legacy Pro to Advanced moves carry a 22 to 38 percent real unit price increase, not the flat renewal the rep is presenting

ServiceNow retired the five legacy tiers on April 9, 2026 and stopped selling old SKUs on July 1, 2026, so every renewal after that date is priced on a structure your expiring contract cannot be compared to directly. The only defensible comparison normalises both sides to the same scope: same fulfiller count, AI stripped to zero, support stripped to zero, consumption valued separately. Until you build that number, you are negotiating against a percentage the vendor chose.

Prepared by Redress Compliance · September 10, 2026 · ServiceNow advisory. Tier reset and renewal engagements, April 2026 to 2026 year end.

The rep will show you two numbers: last year's ACV and next year's ACV, with a percentage between them. That comparison is worthless because the two sides are not measuring the same thing. Your expiring Pro contract bought fulfiller access and nothing else.

The Advanced quote in front of you bought fulfiller access plus a Now Assist entitlement you did not ask for, plus a metered assist pool with an overage rate you have not benchmarked, plus a support tier that repriced itself when Impact Advanced was withdrawn.

Normalisation is four mechanical steps and it takes an afternoon. One: hold fulfiller count constant.

Using the same count on both sides even if you plan to grow, because growth is a separate negotiation and mixing it in is exactly how a 30 percent unit increase disappears into a volume story. Two: strip bundled AI to zero on the new side.

Deducting your own forecast value rather than the list premium. Three: pull Impact out of both sides entirely, because Impact is priced as a percentage of ACV rather than per seat.

So leaving it in means the AI uplift silently inflates your support bill and you count the same increase twice. Four: move the assist pool to its own line valued at the top-up unit rate, not at zero.

What survives those four steps is your true unit price on each side, and in our benchmarking work most legacy Pro to Advanced moves land at a 22 to 38 percent real increase once normalised, against a headline the rep frames as flat or single digit.

The ITOM and CSM row is different and you should treat it as a structural problem, not a negotiation. Foundation was removed for both product lines, so a customer sitting on ITOM Standard has no like-for-like destination and is being pushed up two effective tiers.

There is no argument that recovers that on tier choice alone.

The recovery has to come from unit discount, term length, and a written tier-movement clause, and it should be priced against the benchmark band for your spend tier rather than against your own expiring rate, which no longer describes anything ServiceNow sells.

ServiceNow will value the bundled Now Assist at list: roughly $30 per fulfiller per month for the base premium and around $60 for the Plus tier with agentic workflows and a larger assist allowance.

On a 5,000 seat estate that is $1.8M to $3.6M of "value" handed to you in a slide, and the rep will use it to argue that a 30 percent unit increase is actually a discount.

The worked example the market has settled on is the Pro Plus seat at $175 per user per month, where the AI uplift of 25 to 60 percent adds $44 to $105 per user per month, or $264,000 to $630,000 a year on 5,000 seats. Note what that number is measuring.

It is measuring the uplift applied to your licensed population, not your AI-using population. If 700 fulfillers will realistically touch Now Assist in year one, you are being charged the premium on 5,000.

Build the credit yourself.

Take your forecast of assist consumption by workflow, convert it at the top-up unit rate you can independently benchmark (a case summary runs around 25 assists, a large agentic action around 150), add a 25 percent buffer, and that number is the defensible value of the bundle.

In most first year estimates we see, it lands somewhere between 20 and 45 percent of the list premium, because adoption ramps and the licensed population is far larger than the active one.

Present the credit as a spreadsheet with your own utilisation curve, and read the Now Assist valuation argument before the rep frames the number for you.

The per-seat number is only half the deal. Every new tier bundles a metered assist allowance, and the assist is priced by action complexity, not by user: a case summary burns roughly 25 assists, a large agentic action around 150.

That is a 6x spread between the cheapest and most expensive thing your agents will do all day, and nothing in the per-fulfiller quote reflects it.

Use ServiceNow's own published AI Starter Pack as the only public reference point you have: 25 Pro Plus users at 6,000 assists each, about 150,000 assists total.

Run that against a high-touch ITSM team doing 50-plus AI-assisted resolutions per agent per day and the pool empties in week three of the month, at which point top-up charges apply at a per-unit rate you almost certainly have not benchmarked.

That rate is the second price in the contract, and it is the one with no discount history attached to it, which is exactly why you should be pricing the overage unit before you sign the seat count.

Do three things before the quote is final.

First, demand the Now Assist Overview document by name: it is referenced directly in the Order Form as the authoritative allowance source, it is not on the public price list, and reps routinely quote tier allowances from memory that the document does not support.

Second, forecast consumption from your actual ticket mix (summaries versus agentic resolutions) and add a 25 percent buffer to the initial pool, because buying the buffer up front costs materially less than buying it in-flight. Third, negotiate the top-up unit rate now and cap it for the full term.

ServiceNow will resist pool rollover and will offer a larger year-one pool instead. Take the capped rate over the bigger pool every time.

Read the April 2026 reset for what it actually accomplished commercially.

ServiceNow did not need three AI-native tiers to sell Now Assist; it was already selling Now Assist at roughly $30 per fulfiller per month, and $60 for the Plus variant, as a discrete add-on with a discrete renewal conversation. Folding it into the tier removed that conversation permanently.

What the vendor bought with the repackaging was a new reference price: the number from which every subsequent uplift, every true-up, and every Impact recalculation compounds for the rest of your relationship.

A one-time re-baselining of 25 to 60 percent on the licensed population, applied once and then never revisited, is worth more over a five-year horizon than any amount of add-on attach rate.

This is why the discount percentage is offered so freely. Discount is the one variable that moves the invoice without moving the base.

A rep with authority to concede five points of discount is conceding against a number that has already been reset upward, and the concession is structurally reversible at the next renewal because the list price and the tier definition survive intact.

Ask instead for a three point annual uplift cap and watch the resistance change character: the cap constrains the compounding base, applies to every year of the term, and travels forward into the next agreement as precedent. The vendor knows which of those two asks is expensive. Most buyers do not.

The support line makes the re-baselining self-reinforcing. Impact is priced as a percentage of your Annual Contract Value, so the moment the AI bundle lifts subscription fees, the support fee rises automatically with no negotiation, no signature, and no line item to argue about.

Impact Advanced is gone in 2026, which pushes customers toward the costlier Impact Total, and the increase is routinely presented as a service improvement. You are therefore paying an uplift on an uplift, and the second one arrives without ever being quoted to you.

Now compound that across three years. A buyer who negotiates to a headline discount of, say, 62 percent off list on a re-baselined tier has anchored to a number that says nothing about what a fulfiller costs.

A buyer who normalises both sides to the same scope, AI stripped to zero, support stripped to zero, consumption valued separately, and negotiates to that normalised unit price, has anchored to something that survives contact with the next renewal.

The two approaches can produce identical year-one invoices and materially different year-three ones. That gap is the whole argument for treating unit price rather than discount as the deal point.

Expect the rep to fight the framing rather than the numbers. The standard response is that legacy and new tiers are not comparable, that AI is now inseparable from the platform, and that normalisation is an artificial exercise. Treat that as a tell.

If the tiers genuinely cannot be compared, the vendor cannot claim your renewal is flat either, and the burden shifts back to them to price the delivered scope rather than the packaged scope.

So pick your deal point deliberately and state it once.

A strong outcome looks like a normalised unit price within roughly 5 percent of your expiring effective rate, an uplift cap at 3 percent or lower for the full term, pre-agreed tier-movement percentages in both directions, and a capped assist top-up rate.

Everything else, including the discount headline, is decoration on a base you have already conceded.

Impact is the one line on the quote that cannot be compared per seat at all, because it is not priced per seat.

ServiceNow sets Impact as a percentage of your Annual Contract Value, which means the moment the new tier bundles Now Assist, Workflow Data Fabric, and AI Control Tower into your subscription fee, the support fee rises automatically without anyone quoting you an increase.

That is the compounding mechanic buyers miss: the AI bundle inflates the base, the base inflates the percentage.

And the rep books both as a single "renewal." Add the 2026 withdrawal of Impact Advanced, which pushes a large share of the installed base into the costlier Impact Total and gets narrated as a service upgrade rather than a price event, and you have two increases stacked inside one number.

Expect the rep to resist unbundling on the grounds that Impact is "part of the platform commitment." It is not. It is a separate service with a separate margin, and the vendor sells it separately every day.

The ask is specific and it is not negotiable on your side: subscription and Impact quoted on separate lines.

The Impact percentage fixed in writing for the full term rather than "as applicable," and the percentage applied to pre-AI-bundle subscription value so the support fee does not ride the AI uplift you never asked for.

If ServiceNow will not fix the percentage, cap Impact in absolute dollars with a stated annual growth ceiling. Treat the Impact line the same way you would treat a percentage-of-spend support fee from any other large vendor: the percentage is the negotiation, not the service tier.

Model it before the call: rebuild the Impact fee twice, once on pre-bundle subscription value and once on the quoted new-tier value, and put the delta on the table as its own number.

That delta is real money the vendor has not had to defend, and it is the cheapest concession in the deal because support margin is deep and the sales team is compensated on subscription ACV, not support ACV.

The published bands disagree, and the disagreement is itself useful.

Discount evidence spreads from 15 to 35 percent as the common range, 20 to 35 percent as peer-typical, 25 to 45 percent from roughly 250 fulfillers, 40 to 52 percent above 5,000, and 40 to 70 percent where procurement runs a genuinely competitive process.

Uplift demands cluster at 3 to 10 percent in some sources, 7 to 12 percent in others, and 8 to 12 percent even in accounts where volume grew, which tells you the uplift ask is a policy position rather than a cost-driven number, and an increase alongside growing volume is a negotiable anomaly.

Not a market rate.

On rates, legacy anchors sit at $70 to $100 for Standard, $100 to $150 for Pro, and $150 to $200-plus for Enterprise, while the new three-tier structure is quoted at $70 to $200-plus, which is why the tier label proves nothing.

Volume does the real work: large deployments benchmark at $38 to $90 per fulfiller per month against $100 to $200 for small ones, so the right comparator is your spend tier, not your product tier. Two patterns repeat often enough to plan around.

First, quotes presented as flat or as a modest single-digit uplift routinely normalise to 22 to 38 percent real unit increases once the AI bundle and Impact are stripped out, because the bundle absorbs the difference.

Second, consumption terms, the assist pool size, the overage unit rate, the rollover treatment, are left out of the Order Form and pushed into referenced documentation the vendor can revise.

Both patterns are correctable in the paper, and both are cheaper to fix before signature than at first true-up.

Related advisory: the full topic guide.

How do I compare my legacy ServiceNow Pro rate to the new Advanced tier price?

Hold fulfiller count identical, then strip the bundled Now Assist value out of the Advanced quote before dividing.

Legacy Pro ran $100 to $150 per fulfiller per month and the AI layer is worth $30 to $60 per fulfiller per month, so an Advanced quote at $180 normalises to roughly $120 to $150 on a like-for-like basis.

Anything above your legacy rate after that strip-out is a real increase, regardless of what the discount percentage says.

What is the Now Assist uplift worth in percentage terms?

Between 25 and 60 percent on top of the underlying seat cost, applied to the entire licensed population rather than just the users who actually use AI. On a $175 Pro Plus seat that is $44 to $105 per user per month.

For a 5,000 fulfiller estate that is $264,000 to $630,000 of additional annual cost layered on existing spend.

Can ITOM and CSM customers still get a Foundation tier price?

No. ServiceNow removed Foundation from both ITOM and CSM in the April 2026 redesign, so every ITOM and CSM seat moves to Advanced as a minimum.

That is a structural floor rather than a negotiable tier, so treat the forced uplift as a documented concession you are owed back in discount, uplift cap, or consumption allowance.

Is the bundled AI in the new tiers unlimited?

No. Every tier includes a metered assist allowance with consumption-based overage, and only Virtual Agent conversations are described as unlimited.

A case summary consumes roughly 25 assists while a large agentic action can consume around 150, so high-touch teams running 50 or more AI-assisted resolutions per agent per day can exhaust a monthly pool by week three.

The authoritative allowance numbers sit in the Now Assist Overview document referenced in your Order Form, not in the quote.

Why does Impact support distort the tier price comparison?

Impact is priced as a percentage of your total annual contract value, not as a fixed per-seat charge, so it inflates automatically whenever the AI bundle raises your subscription fees. That means the new tier structure raises your support bill without any separate negotiation.

Strip Impact out of both sides of the comparison, then negotiate the percentage separately and try to fix it against pre-AI-bundle ACV.

What discount should I expect on the new ServiceNow tiers?

Published bands disagree meaningfully. Discounts typically open around 250 fulfillers and reach 25 to 45 percent off list, with 40 to 52 percent reported at 5,000 or more fulfillers and 40 to 70 percent for the most disciplined large-enterprise procurement.

More conservative peer data puts typical outcomes at 15 to 35 percent, so the spread itself is the signal: your leverage, not the tier, sets the number.

Should I negotiate to the discount percentage or the unit price?

Unit price. The discount percentage is measured against a list price ServiceNow controls and reset in April 2026, so conceding on discount costs the vendor very little.

The normalised per-fulfiller-per-month figure and a written uplift cap of 0 to 5 percent are the two numbers that actually govern what you pay across a three year term.