Speaker program compliance, explained.

What a speaker program is, and why a dinner needs a compliance team

From a nomination to a signed report: a speaker program in five steps

1. Plan and nominateThe company decides which topics need programs and hires speakers, on stated criteria and a set fee.
2. Request and approveA rep requests an event: speaker, venue, invitees, estimated meal. Compliance or the bureau approves it.
3. The eventThe talk, the meal and the sign-in sheet. A monitor may attend, ideally without warning.
4. Read what really happenedThe sheet and the invoice are compared with the request. This guide is about this step and the audit choices that follow from it.
5. Decide and keep the recordA monitor clears, changes or refers. Payments go into the yearly federal report, and the file stays.

Step 4 is where most of the evidence first appears. The people and venues in this guide's example are fictional.

Nine suspect characteristics, a payment total back at its old level, and two more surprises in speaker compliance

HCP, honorarium, fair market value, NPI

Fourteen terms are defined in the one-page summary.

How the Anti-Kickback Statute turns a steak dinner into a federal question

In shortThe statute makes it a felony to knowingly and willfully offer, pay, ask for or take anything of value to induce or reward business that federal health programs pay for. A meal counts as value. The speaker, the attendee and the company can each be exposed, and one purpose to win business can be enough.

The Anti-Kickback Statute, 42 U.S.C. 1320a-7b(b), has two halves. One covers whoever solicits or receives remuneration, and the other covers whoever offers or pays it.8 "Remuneration" is the law's word for anything of value, cash or in kind, direct or indirect. The maximum is $100,000 and 10 years in prison, or both.8 A person need not know of this statute or specifically intend to break it.8

Intent is still what the case turns on, and intent is hard to see. That is why the OIG writes that intent "may be evidenced by the speaker program's characteristics and the actual conduct of the parties".1 A compliance file is therefore a record of conduct: who was chosen, who came, what they ate. The OIG's 2003 guidance for drug makers says these arrangements raise the statute "if any one purpose" is to generate business.9

Two other consequences make the exposure large. A claim that includes items "resulting from" a violation counts as a false claim under the False Claims Act,8 the federal law against knowingly making false claims for government money.10 Penalties per claim run $14,308 to $28,619, the range set July 3, 2025, plus three times the damages.11 A conviction also brings mandatory exclusion from federal health programs, which for a drug maker would mean losing Medicare and Medicaid as customers.1

The False Claims Act fight over "resulting from", as of November 2026

Courts disagree on how direct the link must be. In February 2025 the First Circuit, in Regeneron, joined the Sixth and Eighth Circuits in requiring but-for causation for that "resulting from" route: the kickback must be the cause without which the claim would not have been submitted. The Third Circuit asks only for a link. A separate route, false certification, remains available without that proof.12 This narrows one civil route. It leaves the criminal statute untouched, since an offer alone can break it.8

Whistleblower suits, which lawyers call qui tam suits,10 also survived a challenge. On September 1, 2026, the Eleventh Circuit vacated the Florida ruling that they break the Appointments Clause, the Constitution's rule on how officers of the United States are appointed, and sent the other constitutional questions back to the trial court.13 In fiscal 2025 whistleblowers filed 1,297 suits, up from 980 the year before, and False Claims Act recoveries passed $6.8 billion, over $5.7 billion of it from health care.10

Why the OIG doubts the educational value of company-paid speaker programs

In shortThe OIG says it "is skeptical about the educational value" of these programs, because doctors can get the same facts from the label and journals for free. Published studies link industry payments to more prescribing of the paying company's drug. They are observational, so they show association.

The Special Fraud Alert of November 16, 2020 is the document everyone in this field works from. Its argument has two steps. First, investigations found generous pay for talks "offered under circumstances that are not conducive to learning" or to audiences with "no legitimate reason to attend". Second, the same information is available from package inserts, independent conferences and journals, with no money changing hands, which "further suggests that at least one purpose of remuneration" is often to induce or reward referrals.1

The industry's answer, quoted in the alert itself, is that speakers help educate colleagues about benefits, risks and appropriate uses.1 The alert leaves room for that view: lawfulness "depends on the facts and circumstances and intent of the parties", and it says it does not aim to discourage meaningful training.1 Still, it asks companies to "assess the need for in-person programs" and consider less risky ways to convey the information.1

Read these studies for what they are. They show that doctors who receive money prescribe more of the payer's drug, and bigger payments go with bigger effects in 25 studies that looked.14 They cannot show the payment caused the prescribing, since a company may pay the doctors already using its drug. For a compliance team that distinction changes little: the selection itself is what the alert warns about.1

The OIG's nine suspect characteristics, and what a monitor can see before and after

In shortThe alert lists nine characteristics that, alone or together, may point to an unlawful program. Five can be checked from records or the request before the event, one needs a person in the room, and three are confirmed only on the sign-in sheet and invoice. That is why approving the request is half the job.

Wording shortened from pages 5 and 6 of the alert.1 The numbers are ours for reference; the alert's bullets are unnumbered, and the examples are invented for this guide. The alert also says the presence or absence of any one factor "is not determinative".

What an eight-column score of the OIG's nine characteristics gives up

A score can run on eight factors by folding "little or no substantive information" and "no new medical or scientific information" into one column.1 That is a fair working grouping, with a cost: a program can carry fresh information and still present none of it in the room, and only someone watching would know. If your team scores eight, keep the live audit checking content against the approved deck.

OIG, PhRMA Code, AdvaMed and integrity agreements: who says what about speaker programs

In shortThe statute is the only law here. The OIG's alert explains how it sees risk. The PhRMA Code, from the Pharmaceutical Research and Manufacturers of America, is a voluntary standard that companies certify to. A corporate integrity agreement (CIA) is a multi-year compliance contract with the OIG, and it binds only the company that signed it. They differ in useful ways, most of all on restaurants.

Sources: alert,1 PhRMA Code section 7 and question 13,4 Novartis CIA sections III.B and III.L,18 Veloxis CIA section III.L.19

OIG, PhRMA, AdvaMed, Novartis: four status points to get right in November 2026

Several states add their own limits. Vermont, for one, bars most gifts from manufacturers to health care providers, food included, with narrow exceptions, at up to $10,000 per violation.21 Check each state where you hold programs.

Speaker payments in Open Payments: $751.9 million in 2025, back at the level before the alert

In shortOur tally of federal data shows speaker-category payments halved in the pandemic year and have climbed every year since. In 2025 the category held 249,154 payments. Most clinicians receive little. A few thousand receive $50,000 or more from all companies combined.

How we counted. For each program year we added up the total amount of every row coded 2, the category named above, in CMS's "payments grouped by reporting entities and nature of payments" dataset, as published June 30, 2026 and read November 3, 2026, with all recipient types together.3 Dollars are those of each year, with no inflation adjustment. The category can hold other services besides speaking, so it is a ceiling on speaker fees. Non-physician practitioners count from 2021.23

One caution on "level with 2019". Three physician records that year hold $97.7 million from one or two payments each ($16.7 million, $31.0 million and $49.9 million), which look unlike talk fees. Without them 2019 is about $654 million, and 2025 sits above it.22

The OIG used this same category for its "nearly $2 billion" across 2017, 2018 and 2019.1 So the alert's own yardstick is back where it started. Inside that total the mix shifted: teaching hospitals went from $40.7 million to $111.3 million, and individuals received $640.6 million in 2025 against $711.0 million in 2019, or about $613 million with those three records set aside.3,22 The pandemic likely explains 2020: the alert itself noted that in-person activity had been curtailed.1

The per-person numbers add up across all companies a clinician speaks for, while a company's yearly cap applies only to its own payments; PhRMA asks each company to set its cap "individually and independently".4 So a doctor can sit below every company's limit and still draw a large total. Open Payments data is public, and journalists and plaintiffs read it the same way you can. Our tally names nobody and uses the aggregate only.

How a shared speaker dinner lands in Open Payments

Speaker fees go in their own category of the yearly report.24 A shared meal follows a special rule: divide the cost by everyone who ate, office staff included, and report the result only for the people who ate and are covered by the program, such as physicians.24 Items under $13.82 are excluded in 2026 unless a person's total for the year passes $138.13.25

A company that fails to report accurately faces $1,000 to $10,000 per item, capped at $150,000 a year before annual adjustment, and more if the failure is knowing.26 Reports come once a year, so nothing forces a quick look at one dinner.

From Novartis to Takeda: what speaker program settlements reveal about how programs fail

In shortSettlements on speaker programs run from a few million dollars to $900 million, and they arrive years after the conduct. The same patterns repeat in the government's own words. Tying speaker choice to sales or prescribing appears in all six we compared.

How monitoring fell short in the government's telling: advance notice, unnoted meals, a withheld ownership stake

Three details matter for anyone designing a check. At Novartis, audits were announced to reps in advance.27 At Gilead, monitors allegedly did not note meal limits being exceeded or unsuitable venues, so a monitoring program existed and still failed.29 And at Medtronic, which held more than 130 events at a restaurant owned by a neurosurgeon, sales staff allegedly withheld that ownership from compliance. Medtronic paid over $9.2 million and admitted no liability.33

Many of these cases began with an insider. Sales representatives filed at Novartis, Biogen, Biohaven and Assertio, and a physician at Gilead.27,31,30,29,28 Time works against you too: the Gilead conduct ended in 2017 and settled in 2025. What you keep today is what you will have to show in the next decade. A law firm note from August 2026 says the Department of Justice (DOJ) has made clear that drug makers remain a top False Claims Act priority.2

Reconciling a speaker program: per-head cost, attendance counts, NPI checks and pay against the cap

In shortFour small calculations do most of the work of reconciling a program, meaning comparing its sheet and invoice with its request. Divide the bill by the people who really ate. Sort the names on the sheet. Check each NPI. Add the fee to the speaker's total. Here they are on this guide's fictional dinner.

Why a $92 meal per head can really be a $138 meal

The invoice for the example dinner has two lines: dinner for 15 at $980, and a room fee with service and tax at $400. The total is $1,380. Divide by the 15 people on the sheet and you get $92, under the company's $95 limit. Now ask who really ate. Six names on the sheet were not on the RSVP list. If some of them were not at the table, the real figure climbs. On the 10 people invited it is $138.

The room fee deserves its own look. A $400 "room fee, service and tax" on a $980 dinner is 41% on top. Real cases show why that matters. The Gilead settlement carried an allegation that food and drink were booked as room fees to dodge per-person limits.28 Veloxis admitted adding clinicians who never attended to expense reports to lower the per-person figure, at advisory retreats and dinners.6,7 Federal reporting uses the same division: by everyone who ate.24

Sorting 15 signatures into the people you expected

The sheet in the example has 15 rows against 10 RSVPs. One invited doctor did not come, leaving 9. Four doctors walked in without an invitation, and one of them is another paid speaker on the same drug. Two people signed with no NPI: a practice manager and a "guest". Five of the 13 doctors have come to this topic two or more times in 12 months.

A short sheet deserves the same care as a long one. At Insys, the opioid maker whose founder was sentenced to 66 months after a 2019 jury verdict,36 prosecutors alleged that "often no one showed up for the lectures".37 A former rep told PBS Frontline that signatures of other doctors were forged for one dinner; the doctor involved denied any impropriety.38 A sheet that exactly matches its RSVP list can be the suspicious one.

Testing an NPI on the sheet with the Luhn check digit

Every NPI is nine digits plus a check digit, computed with the Luhn formula, the same method payment cards use.39

This helps a lot with handwriting. A single misread digit always breaks the check, so a software reader or a person can flag the row at once. The check proves only that a number is well formed. To learn whose number it is, look it up in the public NPI Registry and compare the name and state. Treat a blank as a finding, and never fill it in with a guess.

Fee, count and cap: adding one more talk to a speaker's year

In the example, Dr. Brandt earns $2,600 a program and has given 12 this year, which is $31,200, or 62% of the company's fictional $50,000 yearly cap. With programs already booked he reaches 73%. PhRMA asks for a yearly cap per speaker and a policy on how many engagements any one speaker should have, and it leaves the figures to each company.4 The 2026 Veloxis agreement requires a centrally managed rate structure from an independent fair-market-value analysis and tracking of total pay to each speaker.19

Seven common speaker program failures, from padded sign-in sheets to speakers taking turns

In shortMost failures come from ordinary incentives: a rep with a budget to spend, a doctor who likes the dinner, a monitor who sees one event in many. None needs a villain, and each leaves a trace in the records if somebody reads them together.

The territory plan, a rep's written plan for a sales region, can be the most direct evidence of all. The alert treats selecting people by "past or expected revenue", including an ROI analysis, as a suspect characteristic.1 Novartis admitted that it measured attendee ROI by new prescriptions.27 A plan or nomination with that language should not be approved, and it will still exist years later if someone ever asks.

How many speaker programs to monitor live: the 10% yardstick, risk-based picks and no advance notice

In shortNo regulation sets a monitoring rate for everyone. We found no published figure for what share of programs companies watch. The best public yardstick is an integrity agreement: 35 programs or 10% a year, whichever is larger, chosen by risk or at random.

Agreements differ. The 2020 Novartis agreement set only four audits per new product or indication, because outside speaker programs were cut back so far.18 Earlier agreements required live audits of up to 125 programs, a law firm's review found.27

The Veloxis text says what a monitor reviews: slides and other materials, speaker statements, and what the sales representative does during the program.19 Those three things cannot be read from an invoice, so records review does not replace a person in the room. The Veloxis case itself concerned meals, retreats and consulting fees, and its agreement sets speaker program rules for the five-year term.19,7

Why the monitor must arrive unannounced

The government faulted Novartis for telling reps about audits ahead of time.27 The law firm Skadden's advice after the Gilead settlement names unannounced live monitoring and randomized audits of program documents.28 PhRMA separately expects a company representative to be physically present at programs, so the rep is in the room either way; what they should not have is a warning.4 Keeping a calendar from the people it watches takes system design: who can see an audit, and who books the extra seat.

Random picks, aimed picks, and what each one gives you

The agreement allows "either a risk-based targeting approach or a random sampling approach".19 Each does a different job. Aimed audits should find more problems per visit, if the ranking works. Random audits give an honest rate of failure, which aimed audits cannot, since they look only where trouble is expected. Random ones also keep reps from guessing the logic. A plan can run both.

Restaurant, hotel, office or webcast: how format changes speaker program risk

In shortThe format sets the risk before anyone sits down. A virtual program has no meal and no venue, and an office talk has a modest one. A restaurant is where the alert's concerns gather, though the industry code still allows some. Companies should write down where they draw the line.

The alert's list names restaurants and "entertainment or sports venues", with examples elsewhere including wineries, stadiums, fishing trips and golf clubs.1 PhRMA bars luxury resorts, high-end restaurants and recreational venues, and says a private room in a local restaurant may suit a talk if the meal is modest.4 Novartis agreed to no restaurants at all.18 One useful test comes from the Code: the venue should not be "the main attraction of the event or perceived as such".4

The speaker program needs assessment: a written answer for each brand, each year

Some companies have gone further and ended programs. Takeda voluntarily ended its Trintellix speaker program, which earned it credit in the settlement, and Pfizer ended Biohaven's after buying the company.5,30 The Novartis model limits outside speakers to the 18 months after a new approval.18 The alert asks every company to weigh the need against "less-risky means".1 Keep a written answer to that question for each brand, each year. The 2026 Veloxis agreement calls this an annual needs assessment and budget, with compliance approving the plan.19

What software can read from speaker program records, and what DOJ will ask about it

In shortSoftware can make coverage of records complete, and it can join sign-in sheets to history in a way no person has time to do. Handwriting remains hard, a score stays a guess until you test it, and judging intent stays with people. DOJ expects you to measure all of this.

We found no published test on speaker sign-in sheets, and those forms are simpler than the benchmark's maternity records: a name, a number, a signature. The benchmark still shows the danger. A reader that invents a value in a blank field can turn a guest with no NPI into a doctor with one. So a good design checks every NPI against its check digit and the registry, leaves blanks blank, and shows the scan beside the reading so a person can compare.

The regulators' questions that do and do not apply to a speaker program score

DOJ's guidance for prosecutors, updated in September 2024, is the one that applies. It asks: "What baseline of human decision-making is used to assess AI?", and how a company measures "the accuracy, precision, or recall of any data analytics models".41 Answer those for your score, with records.

The best-known pharma AI rules mostly do not govern this work.

These records name people. A score next to a doctor's name is an allegation in waiting, so limit who sees it, write down why you hold it, and never let a score stand as a finding by itself. The alert is plain that no single factor decides anything.1

Checklists, samples and spreadsheets: how speaker programs are monitored now

In shortA request is approved against a checklist, a meeting vendor runs the logistics, and the sign-in sheet and invoice wait until someone pulls a quarter's sample. Attendance history lives in a spreadsheet, and live monitors go where the calendar allows. On this guide's fictional estimate below, most of the elapsed time is waiting.

From sales rep to investigator: who touches a speaker program

Spreadsheets, bureau platforms, audit firms: four ways to check speaker programs

A company on an integrity agreement must do more, and in a set form: a centralized electronic system to start and track every program, a full attendee list, sales certifications for each program, and an independent review.19,18 Those requirements are a fair template for any company, since they describe what a regulator wanted to see after something went wrong.

Dr. Brandt's dinner at Bricktown Chophouse: a score of 48 becomes 76 the morning after

In shortThis is the fictional example from the top of the page, inside the product. Lena Ortiz, the monitoring lead, reads the overnight sign-in sheet in one click. Attendance and repeat attendees turn red, the score moves from 48 to 76, and the same speaker's next dinner rises to the top of her unannounced audit picks. Then she signs a report in which every finding is cited.

1

Every program in her region, before the sheet is read

Lena owns seven territories. Each row is a program, and each of the eight columns holds two squares: left for what the request showed, right for what the sign-in sheet and invoice showed afterward. Upcoming programs have a hatched right half, since they have not happened yet.

SP-24106 at Bricktown Chophouse sits at the top with "sheet in" beside it, and it scored 48 from its request: restaurant +18, no new information +8, same topic +10, sales-driven +12.

2

"Read sign-in sheet": two cells turn red

The reader takes the 15 rows and the invoice. Repeat attendees goes from clear to +14: five people have come to this topic two or more times in 12 months, including another paid Dravenzo speaker, Dr. Raymond Kessler, who was not on the RSVP list. Attendance, the flag for people with no business reason to attend, goes to +14 because two people signed with no NPI.

The score moves from 48 to 76, and the line for high risk is 60. The meal stays clear: $92 against a $95 cap, with no alcohol on the invoice.

3

His next dinner rises to the top of the audit picks

Dr. Brandt speaks at the same restaurant again on October 21 (SP-24158). Because his last program turned red at sign-in, and because that venue has now added flags at sign-in, that program's fail odds, a ranking number section 17 explains, rise from 60% to 84%, and it goes to the top of the list.

Lena clicks Plan, and a live audit is booked in her name. The sales force is not told, and the list carries a lock: "The field never sees this list."

4

The program review: six of eight flags, each with its points

The review opens with a dial at 76, the flags and their points, and how each agent scored the program. On the left, the speaker's cap use: 62% now, 73% with programs booked. Below, the attendees, each with their same-topic count.

The decision bar sits at the bottom. After the event, Lena can draft the monitoring report or refer straight to investigations.

5

Every flag cites its line, and she can confirm or dismiss it

Under the dial, each flag states what was found, with numbered citations that open the source. Below each one, a gray line gives the OIG wording and the rule behind it. Lena confirms or dismisses each.

Where the rule is a company's own, the screen says so: "Company rule: review a deck after 250 programs in 12 months".

6

The monitoring report, drafted from eight documents

Lena asks for the draft. It reconciles attendance, lists six findings against policy with their citations, and recommends actions: refer the Kessler and Brandt pairing, move Dr. Brandt's remaining 2026 programs to office or virtual formats, exclude the five repeat attendees from same-topic invitations for six months, re-review the nomination, and ask the brand team to retire or refresh the deck.

The closing line reads "intent is not concluded". She edits, then signs, or signs and refers with the evidence pack to Dana Okafor in investigations.

What the sign-in sheet changed on one dinner, flag by flag

The sign-in sheet, the invoice and the nomination, side by side with the findings

In shortEvery score rests on paper you can open. Every source document sits in one viewer beside the findings: request, attendance list, sign-in sheet, invoice, deck and label, nomination, agreement and territory plan. A citation opens the exact line, so a person can check any claim straight away.

The same review before an event: a wine pairing and a $118 estimate

Sign-in sheets are one half. For SP-24117, Dr. Kessler's dinner at The Corbel Room two days away, the request alone scores 86: a restaurant, $118 a head against the $95 cap, a wine pairing asked for in the request, 6 of 14 invitees who are repeat attendees, and that nomination. Lena can clear it, hold it, refer it, change the venue or send an unannounced audit. On this sample program, moving to an office with no alcohol would take the score to about 52.

Seven agents read and score speaker programs; a compliance monitor decides and signs

In shortEach agent does one narrow job and states what it never does. People decide every program at or above the high-risk line, sign every report, approve every rule change and take every referral. The agents never conclude intent and never contact the sales force.

Lena decides, Marcus and Ama attend, Dana investigates, Priya approves rule changes

Maker and checker stay apart: a monitor's decision and the rule behind it are different people's work, and sales and marketing have no view of the audit plan. Each agent step, decision and signature goes on the audit trail.

A signed report closes the file

How unannounced audit picks work: a formula you can read, a random sample on top and no notice

In shortThe product ranks upcoming programs with a stated formula and tops the plan up with a random sample. The "fail odds" it shows are a ranking number built from that formula, so treat them as a priority order until you have tested them on your own audit results.

How SP-24158 gets to 84: the fail-odds formula, line by line

The sample setting is: 0.8 times the score, plus 16 if the speaker's recent program turned red at sign-in, plus 8 if the venue added flags at sign-in before, plus 4 for four or more repeat invitees, plus 5 if the speaker has not been audited in 30 days, held between 5% and 95%. For SP-24158, scored 70, that gave 60 this morning, then 84 once the sheet from October 6 had been read.

That number is a design choice, and calling it odds is generous until you calibrate it. Random audits give you what you need for that: if programs rated 80% fail about 80% of the time, the number means what it says. The vendor questions in section 23 ask for this, from us as from anyone.

Three weeks of live and virtual audits, kept from the people they watch

The plan puts every audit on a calendar: risk-based picks first, then a random 2% sample of the week's programs, a setting you can change. Two high-risk programs are still uncovered on this morning, both Dr. Kessler's, each one click from an audit. "Advance notices given: 0" is a number on the screen. Reps and the bureau see a monitor arrive, and never a calendar entry.

Two speakers taking turns: what the attendance network and national view show

In shortOne program can look fine while the pattern across programs does not. A network of 12 months of sign-in sheets shows who keeps coming back and which speakers attend each other's talks. A national view shows flags by factor and territory for the chief compliance officer.

What the chief compliance officer sees: flags by factor and territory

What goes into speaker program monitoring, what comes out and what it runs beside

In shortIt reads the records a program already leaves, against your own written rules, and produces a heatmap, cited scores, audit picks and drafted reports. It connects to the systems you already run, keeps roles apart and records every step.

Design targets are the solution's goals, to be measured in your own deployment.

Events platform, HCP master list, deck library: the systems it reads from

It connects by category: the events or meeting platform for requests, RSVPs and costs; a document store or email for scans and invoices; the master list of HCPs for NPI, specialty and city; speaker contracting for nominations, fair market value tiers and pay; the deck library and label history; and the customer relationship management (CRM) system for territory plans.

Referrals can go to your case system with the evidence pack, and anyone with access can ask about a program, speaker or attendee in plain words, and get an answer tied to its source.

Roles, audit trail and security for enterprise compliance teams

Roles keep the audit plan with compliance, give the bureau only change and hold requests, and show the brand lead only saturation by deck. Maker and checker stay separate. Every agent step, decision, signature and rule change goes on an audit trail you can export for an internal auditor, an independent reviewer or a regulator. Monitoring scales to every program across brands and territories, and speaker data stays in the US region. Botminds is SOC 2 Type II audited, and manages information security to ISO 27001.

Those controls support your compliance program. They give you no legal protection under the statute. A product cannot be certified as meeting the Anti-Kickback Statute, and no vendor should claim otherwise. Your rules, decisions and records remain yours to defend.

Meal cap, repeat line and label age: rules you set

What an unread sign-in sheet costs you: settlements at one end, monitor hours at the other

In shortYour biggest exposure is rare and large, and the steady cost is hours spent reconciling paper weeks after anyone can act on it. Public figures frame the first. Nobody publishes the second, so count it yourself in a week.

Timing ten sign-in sheets to find your hours, your delay and your misses

The dinner that ran again, and other costs a speaker program calculator misses

What reading every sign-in sheet returns when you count monitor hours against its yearly cost

In shortPut in your own numbers. The defaults are assumptions, marked as such. The calculator values reconciliation and report-drafting time only. It leaves out the main reason to do this, the bad program you stop before it runs, because no honest formula prices that.

Reconciliation, reports and aim: where speaker monitoring pays

Your hours back across a year of 3,800 speaker programs

If you reconcile only a tenth of your programs by hand today, the time saved falls quickly, and the case rests on coverage. A small program with few reports may not pay on hours alone. That is a fair result: the reason to read every sheet is that you can then act on what it says.

How ready your speaker bureau and compliance team are: records, rules, owners and tests

In shortPick the answer closest to your team today. Eight questions score 0 to 2, and the last is for reflection. The meter adds up as you go, and each question says why it matters and what a strong answer sounds like.

What a score of 0–6, 7–11 or 12–16 asks of your speaker program team

Twelve questions for any speaker program compliance software vendor, including us

In shortAsk these on your own records, with your own monitors in the room. Good vendors will show you the failures as readily as the catches. Ask us the same ones.

Testing speaker program monitoring in ten weeks on one brand and past sign-in sheets

In shortStart with past programs from one brand: you already know how some turned out, nothing is at risk, and a test can include planted errors. Then run new programs in parallel and measure before you widen the scope.

Catch rates, false alarms and days to read: what a speaker program pilot should measure

Speaker program compliance questions, answered: OIG, PhRMA Code, repeat attendees and fair market value

In shortShort answers to what people search for most, each written to stand alone so you can send one to a colleague.