Einstein add-ons are sold as a long list of small line items — Einstein Activity Capture, Einstein Conversation Insights, Einstein Forecasting, Einstein Bots, Einstein Prediction Builder, Einstein Next Best Action, Einstein Discovery, and a growing set of generative capabilities sold as Copilot extensions. Individually, each item looks like a modest decision. Aggregated, they represent a substantial fraction of the total Salesforce bill and a recurring source of cost growth between renewals.
This playbook walks through how to negotiate Einstein add-ons as a coherent portfolio rather than as a stream of point decisions, what concessions to push for at signing and at renewal, and what disciplines distinguish buyers who control their AI envelope from buyers who watch it expand quietly. The playbook reflects patterns we observe across more than 500 engagements where AI add-on costs are now the fastest-growing category of Salesforce spend.
Why add-ons need a portfolio approach
Salesforce sells Einstein add-ons one capability at a time, often through account teams who pursue individual line items rather than the aggregate. The result for buyers is a portfolio of small commitments — each justified at signing, none reviewed in aggregate — that compounds into a meaningful spend line by year two or three. Treating each add-on as a discrete decision misses the portfolio dynamics.
The portfolio approach reframes the conversation. Instead of negotiating each add-on separately, the buyer presents Salesforce with a defined AI envelope — a budget, a scope of use cases, an expected consumption pattern — and negotiates the SKU mix and discount structure inside that envelope. This shifts the discussion from "is this add-on worth $X per user" to "given our AI envelope, what is the optimal allocation between SKUs and consumption."
What to negotiate at signing
At signing, the negotiation focuses on five elements.
SKU portfolio scoped to actual use
License the Einstein capabilities that the deployment plan actually exercises in year one, not the capabilities the account team would prefer to sell. Capabilities that may matter in year two can be added then; pre-purchasing them creates shelfware that the buyer pays for without using. The discipline is to map each Einstein SKU to a named use case with a named owner; SKUs without an owner do not enter the portfolio.
Population right-sizing
Per-user pricing creates per-user cost. Not every user needs every Einstein capability. License each add-on to the user population that uses it, not the user population that has any chance of using it. This right-sizing reduces typical Einstein add-on spend by 25-40% on initial deployments and is among the most consistent sources of negotiated savings we observe.
Consumption envelope
Generative Einstein capabilities are consumption-priced beneath the per-user SKU. The envelope — bundled volume, overage pricing, true-up timing, carryover, reporting — should be in the contract, not inferred from datasheets. Buyers who do not negotiate the envelope explicitly are exposed to mid-term consumption surprises that are notoriously hard to push back on.
Co-terming
Einstein add-ons should co-term with the underlying Sales Cloud or Service Cloud agreement. A separate term creates two negotiation cycles per year and gives Salesforce additional pressure points. Co-terming concentrates the negotiation, preserves the combined-spend leverage, and simplifies renewal planning.
Renewal protections
Negotiate explicit renewal caps on the AI add-ons. The default vendor posture at renewal of an AI add-on is significant uplift; an explicit cap (CPI, or CPI plus a small premium) preserves the deal economics over multiple cycles. Without the cap, the AI portfolio is renegotiated from scratch every renewal — and the buyer's leverage is weaker after the capabilities have been deployed than it was at signing.
What to negotiate at renewal
At renewal, the buyer's leverage on Einstein add-ons depends on the measurement infrastructure built during the term. Three renewal-specific moves matter.
Retirement of underperforming SKUs
SKUs that did not move their target metric should be candidates for retirement at renewal. The discipline is uncomfortable — the account team will push for continued investment with the argument that the deployment "needs more time" — but the data should make the case. Add-ons that demonstrably did not produce value are removed from the portfolio; add-ons that demonstrably did produce value have a stronger case for continued or expanded investment.
Renegotiation of the consumption envelope
The consumption envelope at the original signing reflected forecasts; at renewal it should reflect observed consumption with forward projection. Buyers should bring the consumption history, the value attribution, and the forward forecast as inputs to the renewal — and negotiate the envelope based on actual patterns rather than vendor defaults.
Restructuring the SKU mix
Some Einstein capabilities are best consumed as add-ons; others are best consumed as part of the Einstein 1 Platform bundle; still others may be available through alternative SKU paths. Renewal is the opportunity to restructure the mix based on what the deployment actually does. Buyers locked into a particular SKU structure at signing should use renewal to optimize that structure against operational reality.
The negotiation moves that work
Across the engagements we run on Einstein add-on portfolios, several specific moves consistently produce above-average outcomes.
Bundling at signing. Combining multiple Einstein add-ons into a single negotiated package — with one discount, one consumption envelope, one renewal cap — produces better total pricing than negotiating each separately. Salesforce will price-quote per add-on; the buyer should respond with the package framing and force the conversation to the aggregate.
Pilot scoping. Negotiating an initial pilot scope for a defined user population, with an option to expand at pre-agreed pricing if pilot KPIs are met, protects against overcommitment. The pilot is real adoption; the expansion is optional. Pricing the expansion at signing rather than at expansion negotiation captures the leverage of the initial deal.
Cross-product trading. AI add-ons sit inside a broader Salesforce relationship; concessions on the AI portfolio can be traded against concessions in Sales Cloud, Service Cloud, or Data Cloud. A skilled buyer uses the trades to optimize the relationship economics, not just the individual SKU economics.
Walk-away credibility. The biggest negotiation lever on Einstein add-ons is the credible alternative — Microsoft Copilot, specialist vendors, custom builds on Azure OpenAI. Buyers who articulate the alternative concretely capture more discount than buyers who present the Einstein decision as already made.
What to avoid
Three patterns consistently produce poor outcomes on Einstein add-on negotiations.
Buying the bundle without the use cases. Salesforce sometimes sells the Einstein 1 Platform bundle to organizations whose actual deployment plan exercises a subset of the bundled capabilities. The bundle looks economical at signing and produces shelfware at renewal. Build the bundle when the deployment plan justifies it; buy add-ons when it does not.
Accepting consumption envelopes from datasheets. The datasheet defaults are vendor-favored. The negotiation should produce written, deal-specific entitlements and overage pricing that reflect the buyer's expected consumption pattern, not generic vendor defaults.
Treating AI add-ons as an IT decision. Einstein add-ons depend on user behavior change to produce value. The decision-making should involve the sales, service, or marketing leadership whose teams will operate the capability, not just the IT or platform function. Add-ons purchased without operational sponsorship typically underdeliver.
A negotiation calendar
For organizations approaching an Einstein add-on negotiation or renewal, a calendar of preparation activity produces better outcomes than ad-hoc preparation.
T-minus 12 months: enumerate the existing Einstein add-on portfolio, with adoption metrics, consumption data, and value attribution by SKU. Identify add-ons that should be retired or restructured. Begin internal alignment on the future AI envelope.
T-minus 9 months: develop the forward use-case backlog with named owners and value hypotheses for each. Identify the SKU mix that maps to the backlog. Estimate forward consumption under three adoption scenarios.
T-minus 6 months: develop the credible alternative — vendor, scope, cost — even if not intended for actual selection. Engage alternative vendors in technical discussions to make the alternative concrete.
T-minus 3 months: open the renewal conversation with Salesforce on the buyer's terms. Present the envelope, the backlog, the alternative, and the desired structure. Avoid letting the vendor define the agenda.
T-minus 1 month: finalize the negotiated structure and the contract language. Ensure consumption protections, renewal caps, and co-terming are in writing. Avoid the rushed close that the vendor's quarter-end calendar pressure produces.
Common concessions on signing deals
Across deals where the buyer prepared effectively, common concessions on Einstein add-on signings include:
- Discount — 25-45% off list on the per-user SKUs depending on deal size and competitive pressure.
- Bundled consumption — generative consumption envelope materially larger than datasheet defaults, often 2-3x.
- Overage pricing — per-unit overage pricing 30-50% below list, with renewal-time true-up.
- Renewal cap — explicit cap on renewal uplift, typically CPI or CPI plus 3-5 percentage points.
- Right-to-substitute — option to substitute SKUs within the portfolio at renewal, providing flexibility as the use-case mix evolves.
- Adoption protection — option to reduce per-user volumes within a defined band at renewal based on observed adoption.
- Co-terming — alignment with the underlying Sales Cloud or Service Cloud agreement.
Not every deal achieves every concession. Mid-size deals typically achieve four to five of these; large deals routinely achieve six or seven. The 34% average reduction figure across our portfolio is composed of these concessions in combination — discount alone is rarely sufficient to produce that result without the structural protections that compound it over multiple years.
Closing the negotiation
The Einstein add-on portfolio matters because AI is becoming the fastest-growing line item in many Salesforce relationships. Decisions made at signing about portfolio structure, consumption envelope, and renewal protection compound over multiple cycles. A well-negotiated initial deal produces a manageable AI envelope through years two, three, and four; a poorly-negotiated initial deal produces an envelope that grows faster than the value it delivers and that becomes increasingly difficult to push back on as the program scales.
The work of negotiating the portfolio carefully is a small investment relative to the spend it shapes. Across the engagements we observe, the buyers who invest in the preparation, the portfolio framing, and the structural protections capture materially better outcomes than the buyers who treat Einstein add-ons as individual point decisions. The pattern is consistent across deal sizes, industries, and Salesforce account team configurations — which suggests it is a pattern of buyer-side discipline rather than of vendor-side variation.