MEDDIC vs BANT vs SPIN, checked against what IBM, Huthwaite and the US trademark register actually publish

On 21 April 2026, Chief Judge Wendy Beetlestone granted summary judgment in favour of MEDDICC Ltd and ordered the US Patent and Trademark Office to cancel Registration No. 6,489,058, the MEDDPICC mark. I pulled that registration up on 20 August 2026. It still reads LIVE/REGISTRATION/Issued and Active.

Four months, and no cancellation entry in the file. That gap turns out to be a fair picture of how meddic vs bant vs spin gets covered: the primary sources moved, and the pages explaining them stayed put.

So I counted. On 20 August 2026 I fetched 17 pages ranking for BANT explainers and for this three-way comparison, stripped the tags, and pulled out the sentence where each one expands the acronym. Fourteen say the T stands for Timeline and three say Timing. None of them uses IBM’s own word.

Three things follow. Each takes about a minute to check, and each one contradicts what most of those 17 pages tell you.

What does IBM’s own BANT guidance actually say?

IBM’s PartnerWorld guidance defines BANT as Budget, Authority, Need and Time frame. It treats an opportunity as validated when the prospect meets three of the four items, not all four. And it scopes a BANT-validated response to IBM’s Velocity model, meaning deals under US$100,000, with a separate framework called SID for anything more complicated than that.

The wording is worth reading straight, because there isn’t much of it. On the criteria page IBM writes that “an opportunity is considered validated if the prospect meets three of four of the BANT items”, then allows that “the iSR and sales representative may decide on an either tighter or looser form of BANT”. The four items are listed as budget, authority, need, and “T ime frame = In what time frame will the prospect be implementing a solution?”

Then the part that none of the 17 pages carries. IBM’s lead-validation page says: “A BANT-validated response is suited to, but not limited to, our Velocity (deals less than US$100,000) model. When a deeper level discussion is needed, use SID to validate and qualify opportunities.”

Read that twice. IBM never claimed BANT fitted complex enterprise deals. It built a second framework for them. The received story, that BANT is a legacy enterprise tool which stopped fitting modern deals, has the history running backwards.

IBM also told sellers not to run it as a list at all. Same asset: “Neither BANT nor SID examples are sets of questions for your conversation.” The call-handling page is blunter still, calling it “a flowing conversation, not a scripted call” and reminding reps that “there is no set order in obtaining information”. Which is exactly the behaviour a four-box scorecard produces.

Getting to any of this takes a detour, which is probably part of why only one of the 17 bothered. The live IBM URL 301-redirects to ibm.com/partnerplus, a page carrying no BANT definition anywhere on it, so the readable copy is a Wayback capture. Pipedrive is the one page that still links IBM’s original. Click it and you land on the marketing hub.

What 17 ranking pages say about BANT, checked 20 August 2026

Method, so you can argue with it: I took the URLs from two Google searches run on 20 August 2026, one for the BANT definition and one for this comparison, fetched each with curl and a desktop Chrome user-agent, stripped script and style blocks, then matched the acronym expansion with a single regex looking for Budget, Authority, Need and whatever word followed. Twenty-two URLs went in. Four fell over (three 404s and one 403), and I set aside MEDDIC Academy’s hub because it isn’t a BANT page, which left 17.

The 35,000 calls behind SPIN don’t prove what they get quoted for

The 35,000-call study is the load-bearing citation for SPIN. Neil Rackham spent a chunk of a video published on Huthwaite International’s own site, on 23 April 2026, explaining where it came from.

Asked by Tony Hughes about the size of the sample, he says: “How the hell could a research team watch 35,000 calls in real life? That’s years and years of work.” Researchers didn’t watch them. Huthwaite ran productivity projects at Xerox, IBM and Motorola, took groups of about 20 sales managers, taught them to code a SPIN call, and only let them loose once they hit “a correlation of 0.85 against standard tapes”. Each manager then coached three of their own salespeople across six calls apiece. So the calls were coded by managers who were themselves inside a coaching programme, as part of that programme.

Then the bit that should change how you read the next SPIN summary you meet. Rackham quotes the sceptics back at himself, in his own words: “4 little questions can’t bring about the kind of productivity changes that you’ve got. You’re claiming that the first thousand people who were trained using the SPIN model against control groups had a 17% higher sales volume.” His answer: “it wasn’t the four little questions. It was the whole coaching process, which was also the research process.”

17% against control groups is a real number, and it’s his. It is a number about coaching. Citing 35,000 calls as proof that asking Implication questions in the right order lifts win rates is using a researcher’s data to support a claim he declines to make.

Huthwaite’s live methodology page puts it more politely. The model is built on “4 distinct types of question” which “provide a logical persuasive framework”, it says, “rather than a rigid sequence”. The popular teaching of SPIN as four sequential question types is the one thing the trademark owner says it isn’t.

MEDDPICC’s registered trademark is a claim a US court has now rejected

MEDDICC Ltd sued 01 Consulting LLC in the Eastern District of Pennsylvania, Civil Action No. 24-1836. On 21 April 2026 the court granted summary judgment for MEDDICC and directed the USPTO to cancel the registration. Per the prevailing party’s release, which IPWatchdog also carried, the court found that Lahoutifard “was not involved in MEDDPICC’s conception,” yet “sixteen years after its genesis, he federally registered the term as his trademark”, and that sales professionals “primarily understand MEDDPICC as a framework”. I haven’t read the opinion. Treat the characterisation as one side’s framing and the docket numbers as things you can check.

The register I could check directly. Here is the TSDR status view for serial 88845076 as it stood on the morning of 20 August 2026.

The page stamps its own generation time, “This page was generated by TSDR on 2026-08-20 03:28:42 EDT”, which helps, because this is the most perishable fact here. A register clerk could amend it tomorrow. There may be an appeal running that I know nothing about. All I can honestly report is what the file showed on the morning I looked, and a court ordering cancellation is not the same event as the register recording one.

Both losing-side properties still assert the rights. MEDDIC Academy calls MEDDPICC “a trademark registered with l’INPI, owned by Darius Lahoutifard and used by MEDDIC Academy under exclusive authorization”, which is a French filing the US court did not address. Its sister site goes further, saying “Unauthorized usage of the MEDDPICC name or methodology is a violation of trademark laws”. That was posted in January 2025 and was still live when I read it on 20 August 2026.

The origin dates don’t line up either. MEDDICC.com says “MEDDIC was created in 1996 inside PTC by Dick Dunkel, working under SVP John McMahon alongside teammate Jack Napoli”. MEDDIC Academy says it “originated at PTC in the early 1990s”. MEDDICC’s own April press release describes the acronym as “developed at PTC in the early 1990s”, contradicting its own resource page. Two adversaries in a trademark fight, two incompatible dates between them, and no neutral account of the origin that I could reach.

MEDDIC vs BANT vs SPIN: which one should you run on Tuesday’s call?

Run BANT when the deal is small and you want a quick go or no-go. MEDDIC earns its overhead once a deal is large enough that a single empty field kills it in month four. SPIN sits inside either, because it describes a questioning style rather than a gate, and two of the three were never scorecards at all. Their owners say so in writing.

What changes once you’ve read the owners instead of the summaries is smaller, and more useful on a live call.

Under BANT, change the threshold. Three of four is the rule, so a prospect with a real problem, the right title and a date, but no budget number, is still a validated opportunity by IBM’s own definition. Treating a missing budget as automatic disqualification is the common habit on discovery calls, and it is not what the framework says.

MEDDIC’s payoff lands after the call rather than during it. The value sits in noticing which fields came back empty, and an unfilled Economic Buyer in week two deserves far more of your attention than a tidy Metrics row.

Order is the wrong thing to fix with SPIN. Rackham puts the lift in the coaching loop: baseline call, coded, coached, coded again. If you carry one habit out of SPIN, make it counting your Implication questions against your Situation questions, then getting someone to look at that count with you. That is also the part that survives contact with a prospect who pushes back mid-demo.

If you want that nudge while the call is running rather than in the CRM afterwards, that is what we built Craqly’s Sales Call mode for. Its Discovery sub-mode lists “Qualification framework guidance (BANT, MEDDIC, SPIN)” among what the AI does, and the talk-to-listen gauge targets 35 to 55% rep talk time. Starter is $0 with 20 credits a month, 1 credit being 1 minute of live session, and stealth is on every plan including that one. Pro is $38/mo, or $19/mo billed yearly. Prices checked on craqly.com/pricing on 20 August 2026.

The three frameworks, by what their owners publish

An opinion I could be wrong about. I think MEDDIC’s reputation as the serious grown-up framework is partly a marketing artefact of having two well-funded owners who publish constantly, while BANT’s reputation as the crude one is an artefact of its owner deleting the page. IBM’s three-of-four rule, with a deal-size ceiling and an instruction not to run it as a script, is a more careful piece of sales design than the four-box grid it gets remembered as. Whether that changes anyone’s win rate, I genuinely don’t know.

What I’m not sure about

Two holes in this, and the meddic vs bant vs spin literature shares the first one. I can’t tell you when BANT was invented. Salesforce dates it to the 1950s, Boomerang’s glossary says the 1960s, plenty of pages say 1959, and I could not find an IBM document giving any founding year. The surviving PartnerWorld asset has html_bp_013113 in its URL path, which suggests 31 January 2013, and that dates the page rather than the framework. Every founding year in circulation traces back to somewhere I couldn’t reach, so I’d rather leave a hole than fill it.

The 17-page count is also a sample, not a census, and it was one afternoon’s work. Run the same two searches next month and the numbers will shift, especially if any of those pages read this one. It’s a curl, a tag strip, and one regex, which makes it cheap to reproduce and easy to disagree with.

One last thing worth sitting with. Three companies have already answered this question in public, for free, in their own words. Go read IBM’s archived criteria page instead. It takes four minutes, and it will change what you ask on your next sales call.