Apromore implementation cost: licences, connectors and quotes

Most implementation-cost questions are about moving something into Salesforce. Apromore is the reverse. It reads how work already flows through your org, and through the systems either side of it, and turns that into a process map you can measure, simulate and check for compliance. Nothing is migrated. The cost of an Apromore project is the cost of producing trustworthy event data and acting on what it shows, plus a licence that, as at October 2026, Salesforce sells by quote.

That shapes everything below. If a proposal for Apromore includes a data migration workstream, a field mapping spreadsheet for record loads, or a cutover weekend, it has been written for a different product.

#What you pay Salesforce for Apromore today

Salesforce completed the purchase on 3 November 2025. The announcement framed the product around process intelligence and its own phrase, "agentic process automation", and said: "Salesforce completed its acquisition of Apromore on November 3, 2025." Since then Apromore has kept its own site, now signed "Apromore from Salesforce", and Salesforce's process mining explainer names it directly: "Within Salesforce, Data 360, MuleSoft, and Apromore work together to create a unified view of process execution across systems."

What neither site does, as at October 2026, is publish a price. There is no Apromore line on the Agentforce pricing page, the Sales or Service pricing pages, or a pricing page on apromore.com that we could find. Treat the licence as quote-only and expect the quote to be set by some mix of:

• the number of processes, or process areas, you want to mine
• case and event volume, which grows quickly once Case, Opportunity and Order histories are included
• whether task mining (desktop activity) is in scope as well as system process mining
• which connectors you need, and whether MuleSoft is the integration route
• the number of analysts and business users who need access
• support tier and contract length

Ask your account executive whether Apromore is now quoted as its own subscription or bundled with Agentforce or MuleSoft Automation. The answer changes who signs the order and which renewal it rides on, and the announcement does not settle it.

#The bill around Apromore, line by line

Apromore rarely arrives alone. The surrounding products are where public numbers exist, so price them first and you will know how much of the total is actually the mining licence. All figures below are list prices in US dollars, as at October 2026, from Salesforce's own pricing pages.

The MuleSoft line deserves attention because Apromore gives you a choice. Its Salesforce page states: "You can either use Apromore's dedicated connector for Salesforce to pull data from Salesforce, or MuleSoft's connector for Apromore to push data into Apromore by using MuleSoft as an integration middleware." If you already run MuleSoft, the second route reuses an asset you pay for. If you do not, the dedicated connector avoids buying an integration platform just to feed a mining tool. A partner who defaults to MuleSoft without asking which you own is adding a quote-only licence to your project for their convenience.

Data 360 is similar. Salesforce's own page describes Data 360, MuleSoft and Apromore working together, and for a process that starts in a web form, passes through an ERP and ends on a Case, unifying the events in Data 360 is reasonable. For a process that lives entirely inside Service Cloud, it is optional. Salesforce's pricing page puts it plainly: "Data 360 pricing scales with your usage." So ask the partner to estimate the credits for ingestion and for query separately, using the Data 360 pricing calculator, rather than accepting a single round number. If your org already has Data 360 provisioned for marketing or service work, check how much of the existing credit pool is spoken for before assuming the mining project can ride on it for free.

#What an Apromore implementation is made of

A sound Apromore project has a small number of work packages, and the heaviest one is not where most buyers expect.

1. Process selection. Pick one or two processes with a clear owner and a decision waiting on the answer: lead-to-quote, case resolution, claims handling, order-to-cash. Apromore's Salesforce page lists processes across Service, Field Service, Sales and Financial Services Cloud. A first phase that tries to mine all of them is a discovery exercise with no decision at the end.

2. Event-log design. This is the core. Process mining needs, per process, a case ID that groups the steps, an activity name for each step, and a timestamp. In Salesforce those come from field history tables, status changes, task and activity records, approval history, Flow interviews and sometimes platform events. Someone has to decide which status change counts as "assigned", which as "resolved", and what to do with the case that went from New to Closed in one save. Each of those decisions is a line in the quote and an argument in the workshop.

3. History coverage. Many orgs discover at this point that the history they need was never tracked. Standard field history covers a limited number of fields per object and keeps data for a limited period. Turning tracking on now gives you future data only. Longer retention needs Field Audit Trail, part of Shield, which is a separate quote. Budget a sprint to switch on tracking and let data accumulate before the analysis is credible.

4. Connector build. Either configure Apromore's Salesforce connector or build MuleSoft flows to push events in. The connector route is mostly configuration and credential work. The MuleSoft route is integration development, testing and an extra runtime to monitor.

5. Cross-system stitching. If the process leaves Salesforce, events from the other system need a shared key. That might be an order number, an external ID or a customer identifier resolved in Data 360. This package is where cost can double, because the other system's owners must join the project.

6. Analysis, simulation and conformance rules. Once the log is clean, analysts build the views, define conformance rules (for example, "no refund approved without a manager step") and calibrate simulations. Calibration means checking the simulated cycle times against what really happened before trusting a what-if.

7. Handover to change. The point of mining is a decision: remove a step, add an automation, rewrite an assignment rule, deploy an agent. Apromore's own page names Agentforce, MuleSoft RPA Manager and Flow Orchestration as follow-on targets. That build is a separate project, priced like any other Salesforce delivery, and should be quoted separately.

#What drives the effort up or down

Four things move an Apromore quote more than any others.

The first is history. An org that has tracked status fields for years on the objects that matter is weeks ahead of one that has not. Ask your admin before you ask for quotes.

The second is how many systems a process touches. One system means one extractor. Three systems means three extractors, a key-matching problem and three teams in the room.

The third is how messy your statuses are. If Case Status has been reused for reporting hacks, picklist values renamed, or record types each running their own path, the event-log design takes longer, because each variation must be mapped to a common activity name.

The fourth is the ambition of the first phase. A conformance check on one regulated process can be scoped tightly. A "digital twin of the whole service operation" cannot, and the quote will reflect that, usually as a long discovery line with no fixed end.

There is a quieter fifth driver: who owns the answer. Process mining tends to show that the documented process and the real one differ, and that some of the difference is deliberate. A team that works around an approval step usually has a reason. If nobody with authority is available to decide whether the workaround becomes policy or gets closed, the analysis phase stretches while the findings circulate, and the partner keeps billing for meetings.

Things that reduce effort are equally specific: a single process owner with authority, a recent clean-up of statuses, existing MuleSoft flows that already carry the relevant events, and agreement up front on what decision the analysis will inform.

#Getting quotes you can compare

Send every bidder the same short brief: the process, the objects involved, an estimate of annual case volume, which systems outside Salesforce are touched, whether you own MuleSoft or Data 360 already, and the decision you need to make at the end. Then ask each one to return the quote in the same structure.

Ask for the licence quote from Salesforce separately from the services quote from the partner. Ask the partner to list the events they will extract per process, with the source field or table for each. A partner who cannot name the source of "case assigned" has not scoped the hardest package.

Ask how the connector route was chosen, and what the MuleSoft alternative would cost if they chose the dedicated connector, or the reverse. Ask for a Data 360 credit estimate with the assumptions written out, or a clear statement that Data 360 is not needed.

Finally, ask what the project hands over. A mining project that ends in a slide pack has cost the same as one that ends in a running conformance monitor, but delivered less.

#Costs that are commonly left out

• Field history and Shield. If retention is too short, someone will propose Field Audit Trail late in the project. Price it at the start or rule it out.
• Sandbox capacity. Testing extraction against production-sized volumes needs a Full Copy sandbox. If your contract does not include one, that is an additional purchase.
• Data 360 credits for querying. Ingestion is often estimated. Repeated queries during analysis are often not.
• MuleSoft capacity. If the MuleSoft route is chosen, the flows consume capacity on your MuleSoft contract, which may need to grow at renewal.
• Analyst time. Mining surfaces questions. Someone in the business must answer them, and that time is real cost even when it is not invoiced.
• The follow-on build. Agents, Flow changes or RPA bots that come out of the findings are their own project.
• Privacy review. Event logs carry user IDs and timestamps for staff activity. Task mining in particular can capture desktop activity. Your privacy and employee-relations owners should review scope before a quote is accepted, and their time belongs in the plan.

#Where the other Apromore questions live

The licence question on its own is on Apromore licence cost. How to frame the business question before you start is on Apromore process intelligence. The acquisition itself is on what happened to Apromore, and the connector detail is on integrating Apromore with Salesforce.

If you already run Apromore from before the acquisition and are weighing your options at renewal, read should I migrate from Apromore before signing anything. When you are ready to brief a delivery team, hire an Apromore Salesforce partner lists what to ask them.

The short version for a budget holder: get the Apromore licence quote in writing, price the Data 360 and MuleSoft lines from the public pages or rule them out, and pay for the event-log work properly. That last item is where an Apromore project succeeds or fails, and it is the line most often under-quoted.