An illustrative case · For AI providers selling to pharma
How an AI vendor changed its pharma proposal
Maya, an account lead at an AI software company, has a drafting demo ready for a pharma prospect. Before writing the pilot proposal, she asks the buyer to walk her through a recent brief. What she learns changes the offer.
01 · Discovery
Maya tests the premise behind her demo
The buyer’s brand team wants approved updates sooner when a competitor’s drug is cleared for a new use. These short updates are called competitive briefs.
Key Assumptions Check
“Where did the last brief spend the most time, from request to approval?”
Maya’s drafting demo assumes that faster writing means faster approval. She asks the workflow owner how analysts find sources, check conflicting details, and get a draft reviewed. She requests work logs, analyst notes, and approval timestamps from six recent briefs before putting a scope in front of the sponsor. (U.S. Government, 2009)
02 · Vendor deal review
The vendor decides whether to keep the drafting offer
Back at the vendor, Maya and her solutions engineer use Competing Hypotheses to compare where effort and waiting might arise against the same records. Each explanation would lead to a different proposal. (U.S. Government, 2009)
| Evidence from the buyer | Writing takes most effort | Preparing sources takes most effort | Review adds delay |
|---|---|---|---|
| Work log About three hours per brief went into finding documents and sorting out differences between them. Writing the first draft took about one hour. |
Weakens
Writing took less recorded effort than preparing sources. A drafting pilot would target the smaller task. |
Supports
Finding and checking documents took more recorded effort. The vendor still needs to separate those two tasks. |
Tells us little
The log covers work before the draft goes to review. |
| Analyst notes Analysts gathered many of the same documents again for each brief. |
Tells us little
The notes describe gathering documents, a separate step from writing. |
Supports
Repeated searches add work before drafting. One place to search is worth testing. |
Tells us little
The notes cover work before the draft goes to a reviewer. |
| Approval timestamps Five of six briefs were approved within one working hour of submission. |
Tells us little
These records start when the finished draft is submitted. |
Tells us little
By submission, the team has already gathered and checked the sources. |
Weakens
Most briefs cleared review quickly. The sixth could reveal a review problem worth checking. |
The offer changes from drafting to search
Maya drops the drafting pilot from the proposal. The work log and analyst notes give her a stronger reason to test repeated search. Her solutions engineer checks whether their platform can search a limited set of documents and link each result to its original source. They can scope that test while the buyer confirms which sources it can use.
Quality of Information Check
What Maya checks before promising a result
The log combines search and fact-checking. Maya asks the workflow owner to time them separately. She also asks why the sixth brief took longer to approve. If that brief reveals a recurring wait for a reviewer, she’ll revisit which problem the pilot should address. (U.S. Government, 2009)
03 · Sponsor proposal
Maya brings a narrower offer back to the sponsor
Maya presents a pilot the sponsor can evaluate with the workflow owner and the person who approves data use. The vendor will set up the search and track results with the analysts. The buyer will approve the source list and decide whether the results justify expanding.
The vendor’s initial idea
Write the first draft faster
Pilot a tool that drafts the update.
The vendor’s revised offer
Give analysts one place to search
“Let’s test search on your next three comparable briefs. Your analysts could find the current drug label and the company’s announcement in one place, with a link back to each document.”
- What the vendor asks the buyer to approve
- One workflow, a defined list of permitted sources, and time from the analysts and workflow owner to run the test.
- What both sides measure
- Compare similar briefs under the same review process. Track time spent finding sources, time spent checking differences, and total time from request to approval.
- What would justify expanding
- Approved briefs reach the brand team sooner. Reviewers can still open the original sources and spend no extra time correcting the work.
What if complete briefs were waiting for a reviewer?
If five briefs with complete citations waited two working days for review, Maya would recommend a different first test. The buyer could name a backup reviewer and agree on a response time, then track whether approval gets faster before funding software.
Report 01 works through the questions facing both the provider shaping an offer and the pharma team deciding whether to sponsor it. Explore the report.
From the current issue
Deal Formation
From Interest to a Pilot Decision
Written for sponsors and AI providers, Report 01 follows a commercial partnership from initial interest through the pilot decision. Its examples focus on U.S. pharmaceutical commercial operations.
48-page PDF. Eight structured analytic techniques. $99 for one named reader.
The continuing series
Put better questions to work.
Subscribe to Deal Design Intelligence for practical cases, analysis and new-report announcements. Start with practical cases on the decisions behind enterprise AI partnerships in life sciences.
New analysis as it is published. Full reports are sold separately.
Strategic areas are optional. Leave them blank for the current enterprise AI partnership series, or select only the areas you want. Broader areas are proposed future coverage.
The sign-up form is loading.
Read how we handle information.