TL;DR
Ad board decks are discussion instruments, not presentations
Content density is deliberately lower, and every stimulus slide exists to provoke a specific response from the panel - not to inform them of something they already know better than you do.
The deliverable is three artefacts, not one
The pre-read circulated ahead of the meeting, the live discussion deck used in the room, and the synthesis capture template that turns four hours of expert opinion into a structured output.
The failure mode is almost always over-production
A polished 60-slide data review, and no capture framework at all.
Turnaround is the binding constraint
Protocol changes, late data cuts and last-minute KOL substitutions mean the final deck is often locked the night before the meeting. Overnight production is not a premium feature here; it is the baseline requirement.
An advisory board deck has one job that no other pharmaceutical presentation shares: it has to get eight to fifteen external experts talking, disagreeing, and going on record - and then capture what they said in a form the brand team can act on by Monday. It is not a deck that presents a conclusion. It is an instrument for extracting one. Teams that brief it like a congress symposium or a launch deck get a beautiful set of slides that runs out of things to do forty minutes into a four-hour meeting.
The most common failure I see in advisory board materials is a deck built to demonstrate the sponsor's competence rather than to elicit the panel's. If your first substantive slide is a twenty-minute data review of evidence your advisors helped generate, you have spent the most valuable hour of the meeting telling experts what they already know. The strongest ad board decks I have worked on open the discussion within fifteen minutes and treat every subsequent slide as a question with a visual attached. Design that in from the brief, or you will not recover it in the room.
- Vaidehi Shukla, Principal, Life Sciences & Market Research, A1 Slides
A launch deck, an MSL presentation, or a payer submission carries an argument from opening to conclusion. An advisory board deck carries a sequence of questions. The structural unit is not the section - it is the discussion block: a short stimulus, a framed question, and enough white space on the slide that the facilitator can annotate live without the layout collapsing.
This changes concrete design decisions. Slides carry fewer data points, deliberately. A forest plot that would be fully annotated in an MSL deck is stripped back to the comparison that the question actually turns on, because a slide carrying six competing signals invites the panel to litigate the data rather than answer the question. Action titles behave differently too: instead of stating the finding, the title states the question under discussion. "Which subgroup should define the target population?" earns its place on an ad board slide. "Subgroup A shows superior response" does not - it pre-empts the panel and biases the output you are paying to collect.
Content density is a compliance question as much as an editorial one. Advisory boards exist to gather advice, not to promote. A deck that reads as promotional creates real exposure, and it changes how the meeting is characterised. The FDA's advisory committee overview is a useful external reference point for how expert advice is separated from promotional communication. The design language should read as scientific exchange throughout - neutral colour treatment, no product imagery, no commercial claim architecture. Our broader approach to that visual grammar is set out in the clinical data visualization playbook .
Most ad boards now run at least one structured elicitation exercise: ranking, voting, or scoring. Whether the mechanism is a polling platform, an anonymous handset system, or a facilitator marking up a slide in real time, the slide has to be built to be modified while thirty people watch.
Practically, this means every polling slide ships with its result state pre-built - the empty bar chart, the ranking frame, the scored matrix - so the facilitator populates a designed asset rather than typing into a text box on a live screen. It means editable native objects, never flattened images, for anything that will be annotated. It means an alternate high-contrast version of any slide that will be projected in a hotel meeting room with the lights up, because a colour ramp calibrated for a laptop screen will not survive that environment. And for hybrid boards, it means checking legibility at the smallest tile size a remote advisor will actually see.
The pre-read is a dense, referenced, self-contained document. It carries the evidence base, the therapeutic context, the disclosure of what is published versus data on file, and the agenda. Advisors read it alone, at speed, usually on a flight. It should be built for reading.
The live deck is the opposite: sparse, high-contrast, built for a room. The single most common production error is shipping the same file for both, which produces a pre-read too thin to prepare from and a live deck too dense to run a discussion against.
The split also matters for approvals. The pre-read typically carries the heavier review burden because it circulates externally and unaccompanied. Building them as one file means the whole package moves at the speed of the slowest review - which is how a deck ends up locked at midnight before an 8am start.
Agenda with timings, advisor biographies, therapeutic and competitive context, the evidence base with full referencing, explicit labelling of published versus data-on-file material, and the specific questions the board is being asked to address. Circulated far enough ahead to be read.
Three to five slides establishing scope, the decisions the sponsor faces, and - critically - what the board is not being asked to do. Boards that drift do so because scope was never bounded on a slide.
The working core. One question per slide, one visual per question, generous margins for live annotation. Built as a modular sequence so blocks can be reordered or dropped when a session runs long, which it will.
Pre-built result states, editable objects, a designed empty frame ready to be filled. Anonymous mechanisms where the question is one advisors would not answer candidly in front of peers.
The most under-produced asset in the entire category. Structured frames - theme grids, consensus-versus-divergence tables, prioritisation matrices - that a scribe populates during the meeting rather than reconstructing from notes afterwards. Where a board produces a same-day readout, this is what makes it possible.
Indexed, searchable, and organised by anticipated question. When an advisor asks for the Kaplan-Meier curve by prior line of therapy, the facilitator finds it in ten seconds or the moment passes.
A pre-built summary template aligned to the capture frames, so the medical affairs team is populating a designed document rather than starting a new one at 11pm.
The most expensive mistake, and the most common. A team briefs a generalist agency, the agency produces a comprehensive, well-designed presentation of the evidence, and the meeting becomes a lecture delivered to people who wrote the papers. The panel disengages, the sponsor's own team fills the silence, and the output is thin. The diagnostic question at brief stage: what does the slide ask the advisors to do? If the answer is "understand this," it belongs in the pre-read.
Companies invest heavily in the input deck and nothing in the output. Advice is captured in a scribe's freeform notes, then reconstructed into a summary days later - by which point the specific language of a key recommendation has softened into a general theme. For a meeting carrying honoraria, travel, venue and internal time, the marginal cost of building capture templates alongside the deck is negligible and the return is immediate.
A generalist design partner will cheerfully add a competitor comparison, a persuasive superiority claim, or an unlabelled data-on-file figure, because those are normal moves in commercial design. In an advisory board context they are liabilities. A partner working in this category needs to know without being told: unapproved indications are handled as scientific exchange, every figure is traceable to a source, data on file is labelled as such, and the deck's register stays non-promotional throughout.
Overnight, against the meeting date. Briefs submitted in your evening return finished in your morning. That cadence is the operating model, not an escalation path - which is what makes it usable when a data cut lands three days before the board convenes and the discussion slides all need rebuilding.
NDA-secured handling of unpublished data. Advisory board materials routinely contain unpublished readouts, competitive intelligence and pre-approval positioning. Every engagement runs under binding NDA with access-controlled environments and secure transfer, with no external storage. Confidential source data does not leave the controlled workflow.
MLR-aware construction . Referencing and annotation layers are built in rather than retrofitted, figures are traceable, data-on-file material is labelled at build, and slides ship in review-ready form. We do not perform your MLR review - but we do not hand you a deck that fails it on formatting grounds either.
Specialist POD, not a general queue. Life sciences work runs through a dedicated POD trained on therapy-area conventions and the visual language of medical affairs. Content-sensitive work is reviewed under the oversight of our Life Sciences principal .
Full-package production. Pre-read, live deck, polling states, capture templates and output shell, produced as one coherent system in a single brand architecture - which is how a same-day readout stops being aspirational. For late-stage production pressure, see our note on overnight presentation design .
The pre-read should be locked and circulated far enough in advance for advisors to actually read it; the live deck can and often should be finalised much later, because late data and agenda shifts are normal. Building the two as separate files is what makes the split possible - otherwise a change to one holds the other hostage.
Yes - that is the standard case, not the exception. A messy working file, a Word agenda and a set of source figures is enough to start. Where the discussion architecture is still unsettled, we build the modular blocks first so sections can be reordered without a rebuild.
Under binding NDA, in access-controlled environments with secure transfer and no external storage. Data-on-file and unpublished material is labelled as such at build so the distinction is visible to reviewers and advisors rather than assumed.
Both, and we recommend briefing them together. Capture frames designed after the discussion deck rarely align to the questions actually asked. Built alongside, each discussion block ships with the frame that will hold its output - which is what allows a readout to be drafted during the meeting rather than reconstructed from it.
A1 Slides helps global enterprises transform data, strategy, and research into PowerPoint presentations that accelerate clarity and decision-making. Our Insight-First Design methodology powers communication for Fortune 500 teams, consulting firms, and industry leaders worldwide.
Read More