Latest / Uphoff on Media Podcast / The Three Decisions a Machine Can’t Make for You

Transcript

The Three Decisions a Machine Can’t Make for You · Uphoff on Media Podcast

• 0:02
  SPEAKER_00: Hey there, it's Tony.
• 0:03
  SPEAKER_00: Over 35 years of running B2B businesses, I keep coming back to the same three decisions.
• 0:10
  SPEAKER_00: Who do we serve?
• 0:12
  SPEAKER_00: What do we sell them?
• 0:13
  SPEAKER_00: And what is it worth?
• 0:15
  SPEAKER_00: Positioning, packaging, and pricing.
• 0:17
  SPEAKER_00: Get them right and almost everything downstream gets easier.
• 0:20
  SPEAKER_00: Get them wrong and no amount of sales effort fully makes up for it.
• 0:25
  SPEAKER_00: For most of my career though, you could get away with being a little vague on all three.
• 0:30
  SPEAKER_00: A good sales team covered for you.
• 0:32
  SPEAKER_00: Your best rep would sit across from a buyer,
• 0:34
  SPEAKER_00: translate the homepage into the buyer's problem,
• 0:38
  SPEAKER_00: pull together a bundle on the call and find a price that buyer could take
• 0:42
  SPEAKER_00: to finance. The alignment happened live at the point of contact.
• 0:47
  SPEAKER_00: That's changing.
• 0:48
  SPEAKER_00: Every B2B buying committee now has a new member.
• 0:51
  SPEAKER_00: An AI agent that does the early research,
• 0:54
  SPEAKER_00: compares you with every alternative and builds the shortlist
• 0:57
  SPEAKER_00: before your sales team even knows a deal exists.
• 1:01
  SPEAKER_00: It reads exactly what you publish and it doesn't wait for your best rep.
• 1:05
  SPEAKER_00: If your positioning, packaging or pricing doesn't hold up on its own,
• 1:09
  SPEAKER_00: you lose the deal before it starts and you never find out it was there.
• 1:14
  SPEAKER_00: So today we're gonna work through those three decisions,
• 1:17
  SPEAKER_00: why they matter more in the agentic era than they ever have
• 1:20
  SPEAKER_00: and why they've never been harder to get right.
• 1:23
  SPEAKER_00: I'll also walk you through something I built to deal with this, buying committee modeling.
• 1:28
  SPEAKER_00: It puts an offer in front of an AI model of every seat on the committee plus the buyer's agent.
• 1:33
  SPEAKER_00: It doesn't predict who will buy.
• 1:34
  SPEAKER_00: It shows you where your positioning, packaging, and pricing are working,
• 1:38
  SPEAKER_00: where they aren't, and which customers your offer actually fits.
• 1:43
  SPEAKER_00: As always, the companion post on Uphoff on Media has
• 1:46
  SPEAKER_00: an Uphoff advisory framework you can download and use in your business.
• 1:50
  SPEAKER_00: This one is the three Ps, three questions worked in order.
• 1:55
  SPEAKER_00: Okay, let's get to work.
• 1:59
  SPEAKER_00: Go to Intercom's pricing page and you'll find
• 2:02
  SPEAKER_00: the same company selling customer service two different ways.
• 2:06
  SPEAKER_00: Human agents are priced the old way, $29 to $132 per seat per month.
• 2:13
  SPEAKER_00: Its AI agent, FIN, costs $0.99 per outcome with no seats required.
• 2:20
  SPEAKER_00: That page is a small picture of a large shift.
• 2:23
  SPEAKER_00: When software does the work,
• 2:25
  SPEAKER_00: the seat stops measuring the value and the price has to find a new unit.
• 2:30
  SPEAKER_00: The number on the page is the easy part.
• 2:32
  SPEAKER_00: The unit underneath it is where the business gets decided.
• 2:37
  SPEAKER_00: Here's the pattern I see across B2B information, marketing and technology companies right now.
• 2:42
  SPEAKER_00: The cost of building a product is falling and the cost of copying one is falling faster.
• 2:47
  SPEAKER_00: In the businesses I work with,
• 2:49
  SPEAKER_00: a feature that took a competitor a year to copy now takes about a quarter.
• 2:54
  SPEAKER_00: When the product gets cheap to make and easy to compare, the value doesn't disappear.
• 2:58
  SPEAKER_00: It moves into three decisions no machine can make for you.
• 3:03
  SPEAKER_00: What you stand for, what you bundle, and what you charge.
• 3:06
  SPEAKER_00: Positioning, packaging, pricing.
• 3:10
  SPEAKER_00: I've used those three to diagnose businesses for years.
• 3:12
  SPEAKER_00: The ones I've led and the ones I advise.
• 3:17
  SPEAKER_00: The buying committee has a new member.
• 3:19
  SPEAKER_00: Every B2B company now sells into a buying committee and to the machine
• 3:23
  SPEAKER_00: that committee sends ahead.
• 3:25
  SPEAKER_00: The data on the second buyer is no longer thin.
• 3:28
  SPEAKER_00: In a Gartner survey of 646 B2B buyers published in March,
• 3:34
  SPEAKER_00: 67% said they prefer a buying experience with no sales rep and 45% said they used AI
• 3:40
  SPEAKER_00: during a recent purchase.
• 3:45
  SPEAKER_00: A 2025 buyer experience report drawn from more than 4,000 buyers found
• 3:50
  SPEAKER_00: that 94% use LLMs, large language models,
• 3:55
  SPEAKER_00: during the buying process and that the vendor a buyer favors before talking
• 3:59
  SPEAKER_00: to anyone still wins 80% of the time.
• 4:03
  SPEAKER_00: Read those two findings together.
• 4:05
  SPEAKER_00: The shortlist is being built before your team knows a deal exists and the vendor
• 4:09
  SPEAKER_00: that leads the shortlist usually wins.
• 4:11
  SPEAKER_00: The agent doing the early research applies a pass or fail test
• 4:14
  SPEAKER_00: to whatever it can read about you.
• 4:16
  SPEAKER_00: The committee that follows applies a different test.
• 4:19
  SPEAKER_00: Can I defend this choice to my CFO and my board?
• 4:23
  SPEAKER_00: Positioning, packaging and pricing are how you pass both tests.
• 4:27
  SPEAKER_00: They are also where most B2B companies are weakest and for years that weakness was affordable.
• 4:33
  SPEAKER_00: Sales and Marketing did the alignment work live at the point of contact.
• 4:36
  SPEAKER_00: A good rep translated a vague homepage into the buyer's problem,
• 4:41
  SPEAKER_00: built a bundle on the call, and found a price the buyer could take to finance.
• 4:45
  SPEAKER_00: The three decisions didn't have to line up on paper because a person lined them up in the room.
• 4:50
  SPEAKER_00: An agent doesn't wait for that conversation.
• 4:53
  SPEAKER_00: It reads what you publish, compares you with every alternative,
• 4:56
  SPEAKER_00: and passes or fails you before anyone on your team knows a deal even exists.
• 5:01
  SPEAKER_00: Miss on any of the three and you lose the deal before it starts
• 5:04
  SPEAKER_00: without ever learning it was there.
• 5:06
  SPEAKER_00: That is why the cost of getting these decisions wrong has gone up.
• 5:10
  SPEAKER_00: The sales team is no longer in the room when the shortlist gets made.
• 5:15
  SPEAKER_00: Putting the offer in front of the room.
• 5:17
  SPEAKER_00: So I've been building that room.
• 5:19
  SPEAKER_00: I call it buying committee modeling, and it is now an Uphoff advisory solution for clients.
• 5:25
  SPEAKER_00: Here's how it works.
• 5:26
  SPEAKER_00: A typical model seats seven or eight roles from the executive sponsor to finance
• 5:31
  SPEAKER_00: to the team that will use the product,
• 5:33
  SPEAKER_00: each an AI persona built around how that role weighs a purchase.
• 5:38
  SPEAKER_00: A separate agent stands in for the buyer's AI running the same searches it would use
• 5:42
  SPEAKER_00: to build a shortlist.
• 5:44
  SPEAKER_00: Then the offer goes in front of them, the position, the package,
• 5:47
  SPEAKER_00: and the price as a buyer would find them in two to four different buying situations.
• 5:53
  SPEAKER_00: Each seat answers on its own three times in each situation,
• 5:57
  SPEAKER_00: so a single offer produces dozens to more than 100 independent responses.
• 6:04
  SPEAKER_00: Every answer follows the same format.
• 6:06
  SPEAKER_00: How likely the seat is to take a meeting and to shortlist the offer?
• 6:10
  SPEAKER_00: A stance from champion to blocker, the objection that would kill the deal,
• 6:14
  SPEAKER_00: and what would turn it into a yes?
• 6:16
  SPEAKER_00: Those answers are tabulated as medians and ranges, never as a single opinion.
• 6:22
  SPEAKER_00: Each committee then deliberates to a verdict, advance, advance with conditions,
• 6:26
  SPEAKER_00: park or decline.
• 6:28
  SPEAKER_00: Finally, a blind reviewer that never sees the committee's answers scores the offer
• 6:32
  SPEAKER_00: on seven dimensions from value clarity to price to value to how easily
• 6:36
  SPEAKER_00: an agent can read it and names the one to fix first.
• 6:40
  SPEAKER_00: What comes back is the conversation your sales team never gets to hear,
• 6:44
  SPEAKER_00: the objections, the questions finance will ask,
• 6:47
  SPEAKER_00: and whether the agent finds you at all.
• 6:49
  SPEAKER_00: The modeling doesn't predict who will buy.
• 6:51
  SPEAKER_00: It guides the three decisions, showing where your positioning, packaging,
• 6:55
  SPEAKER_00: and pricing are working and where they aren't before a real buyer sees them.
• 6:59
  SPEAKER_00: It also validates your ideal customer profile.
• 7:02
  SPEAKER_00: Put the same offer in front of different committees, and you see which buyers it actually fits.
• 7:07
  SPEAKER_00: And you can run it again every time you change one of the three decisions.
• 7:13
  SPEAKER_00: I started with my own firm, and I now use it with clients to test positioning,
• 7:17
  SPEAKER_00: packaging, and pricing before they go to market.
• 7:19
  SPEAKER_00: I'm not the first to build something like this,
• 7:21
  SPEAKER_00: but very few companies have seen their offer the way this room sees it.
• 7:26
  SPEAKER_00: I'll share what the modeling is showing in future posts.
• 7:32
  SPEAKER_00: Positioning has to survive being read literally.
• 7:35
  SPEAKER_00: An agent reads what you publish, not what your best rep says on a call.
• 7:39
  SPEAKER_00: Quote, solutions that accelerate growth, end quote,
• 7:43
  SPEAKER_00: passes no test because it describes every company in the category.
• 7:47
  SPEAKER_00: Positioning in the agentic era has to name the buyer, the problem,
• 7:50
  SPEAKER_00: and the result in words specific enough that a machine can match them to a request
• 7:55
  SPEAKER_00: and a human can repeat them in a meeting.
• 7:57
  SPEAKER_00: That is a higher bar than most homepages clear.
• 8:00
  SPEAKER_00: It's also the first decision because the other two follow from it.
• 8:04
  SPEAKER_00: You can't package around a promise you haven't made and you can't price
• 8:07
  SPEAKER_00: a result you haven't claimed.
• 8:09
  SPEAKER_00: When I was brought in to transform ThomasNet, the work started here.
• 8:12
  SPEAKER_00: The company was known as an industrial directory.
• 8:16
  SPEAKER_00: The value buyers actually got was sourcing,
• 8:18
  SPEAKER_00: finding a qualified supplier faster with data they could trust.
• 8:22
  SPEAKER_00: Repositioning around that job and rebuilding the business as a sourcing
• 8:26
  SPEAKER_00: and data platform is what made the rest of the transformation possible.
• 8:30
  SPEAKER_00: Packaging is where most of the money hides.
• 8:33
  SPEAKER_00: In nearly every company I've been brought in to turn around or transform,
• 8:36
  SPEAKER_00: I found the exact same thing.
• 8:39
  SPEAKER_00: Products sold a la carte to customers who wanted an integrated solution
• 8:42
  SPEAKER_00: that worked like a recipe.
• 8:44
  SPEAKER_00: The catalog described what the company made.
• 8:46
  SPEAKER_00: It didn't describe what the customer is trying to get done.
• 8:49
  SPEAKER_00: At ThomasNet, moving from a la carte to solution pricing raised the average sale by nearly 40%.
• 8:55
  SPEAKER_00: At Pipeline360, we built a branded demand solution
• 8:58
  SPEAKER_00: that combined programmatic display with data-driven demand generation at one price.
• 9:03
  SPEAKER_00: Order size and retention went up and campaign performance improved by more than 35% on average.
• 9:09
  SPEAKER_00: The products didn't change much, the packaging did.
• 9:13
  SPEAKER_00: Agents raise the stakes on this.
• 9:15
  SPEAKER_00: A buyer agent comparing vendors breaks an a la carte catalog into line items
• 9:19
  SPEAKER_00: and prices each one against the market,
• 9:21
  SPEAKER_00: which is how a product gets turned into a commodity.
• 9:24
  SPEAKER_00: A program built around an outcome is much harder to pull apart.
• 9:28
  SPEAKER_00: For digital media and marketing products, the test I use is simple.
• 9:32
  SPEAKER_00: An ad, sponsorship, or lead generation product earns its place on the rate card
• 9:36
  SPEAKER_00: when it sells at least half of its available inventory.
• 9:39
  SPEAKER_00: Below that, fold it into a program or retire it.
• 9:42
  SPEAKER_00: Most media companies I review carry several products that fail that test
• 9:46
  SPEAKER_00: that are still being sold, produced, billed, and reported on every month.
• 9:52
  SPEAKER_00: Pricing is breaking at the unit, not the number.
• 9:55
  SPEAKER_00: Seat-based pricing assumes a human does the work and the software helps.
• 9:59
  SPEAKER_00: When an agent does the work,
• 10:01
  SPEAKER_00: the number of seats shrinks while the value delivered goes up and a price built
• 10:05
  SPEAKER_00: on seats falls exactly when the product is getting better.
• 10:08
  SPEAKER_00: That's the reason a company like Intercom now prices the same work two ways on one page.
• 10:14
  SPEAKER_00: The fix is a new unit chosen from what the buyer counts as value, an outcome,
• 10:18
  SPEAKER_00: a resolved case, a qualified lead, a decision supported.
• 10:22
  SPEAKER_00: Picking that unit is a positioning decision as much as a pricing one,
• 10:26
  SPEAKER_00: which is why the three can't be worked in isolation.
• 10:29
  SPEAKER_00: The discipline underneath doesn't change.
• 10:32
  SPEAKER_00: As I argued in the post, thanks for the lecture on price,
• 10:35
  SPEAKER_00: a price is the last step of a process that starts with the market and the financials
• 10:39
  SPEAKER_00: and ends with the price to value case a buyer can defend.
• 10:43
  SPEAKER_00: What has changed is how quickly a weak case gets exposed.
• 10:48
  SPEAKER_00: Why it's never been harder.
• 10:50
  SPEAKER_00: There are four reasons and they compound.
• 10:52
  SPEAKER_00: Two buyers apply two different tests.
• 10:55
  SPEAKER_00: The agent wants literal, comparable facts.
• 10:57
  SPEAKER_00: The committee wants a story it can defend.
• 11:00
  SPEAKER_00: Positioning that is precise enough for the first and compelling enough
• 11:03
  SPEAKER_00: for the second takes more work than either one alone.
• 11:07
  SPEAKER_00: The three decisions are now tied together.
• 11:09
  SPEAKER_00: Change the pricing unit to outcomes and the package has to change to deliver them.
• 11:13
  SPEAKER_00: Promise an outcome in your positioning and you have to be able to measure it
• 11:15
  SPEAKER_00: or the price falls apart in the first renewal.
• 11:19
  SPEAKER_00: Copying is faster.
• 11:20
  SPEAKER_00: A competitor can match your package in a quarter and undercut your price in a week.
• 11:25
  SPEAKER_00: The advantage now goes to whoever revisits the three decisions most often.
• 11:30
  SPEAKER_00: Outcome pricing needs data most companies don't have.
• 11:33
  SPEAKER_00: You can't charge per result if you've never measured results.
• 11:36
  SPEAKER_00: For many B2B companies, the first year of outcome pricing is really the first year
• 11:40
  SPEAKER_00: of measuring what they actually delivered.
• 11:46
  SPEAKER_00: Here's an Uphoff advisory framework you can put to work on any offer you sell.
• 11:51
  SPEAKER_00: I call it the three Ps, positioning, packaging, and pricing.
• 11:54
  SPEAKER_00: Three decisions work in order, and the order is the point.
• 11:57
  SPEAKER_00: Never package before you position, never price before you package.
• 12:01
  SPEAKER_00: For each one, there's a question to answer.
• 12:03
  SPEAKER_00: Two tests and a warning sign.
• 12:05
  SPEAKER_00: The agent tests, asks whether a machine can read you and compare you.
• 12:09
  SPEAKER_00: The committee tests, asks whether people can defend you.
• 12:12
  SPEAKER_00: Start with positioning.
• 12:13
  SPEAKER_00: The questions, who do we serve?
• 12:16
  SPEAKER_00: What products do we own and what result do we deliver?
• 12:19
  SPEAKER_00: The agent test, can an agent match your words to a buyer's request using only what you publish?
• 12:25
  SPEAKER_00: The committee test, can a buyer repeat your positioning in one sentence to their CFO?
• 12:29
  SPEAKER_00: The warning sign, your homepage could describe any company in your category.
• 12:34
  SPEAKER_00: Second, packaging.
• 12:35
  SPEAKER_00: The question, how does this offer fit with the rest of what you sell and with what
• 12:40
  SPEAKER_00: the customer already uses to deliver outcomes they can measure?
• 12:43
  SPEAKER_00: The agent test, does it hold together when an agent prices each part against the market?
• 12:48
  SPEAKER_00: The committee test, does it read as a program the buyer can defend or as a catalog?
• 12:53
  SPEAKER_00: The warning sign, products that sell less than half their inventory are still on the menu.
• 12:59
  SPEAKER_00: Third, pricing.
• 13:00
  SPEAKER_00: The question, what does the buyer count as value and how do your unit economics hold
• 13:05
  SPEAKER_00: that at that price?
• 13:06
  SPEAKER_00: The agent test, is the price published plain and comparable?
• 13:09
  SPEAKER_00: The committee test, is there a price to value case a CFO can model?
• 13:13
  SPEAKER_00: The warning sign, you price per seat while your product does more of the work.
• 13:18
  SPEAKER_00: If you spot a warning sign, fix the earliest one first.
• 13:22
  SPEAKER_00: The full chart is in the post or at upoffadvisory.com slash frameworks.
• 13:31
  SPEAKER_00: The decisions you leave unmade get made for you.
• 13:34
  SPEAKER_00: Your competitors will be able to copy most of what you build by next spring.
• 13:38
  SPEAKER_00: What they can copy is a clear answer to who you serve, what you sell them, and what it's worth.
• 13:43
  SPEAKER_00: If you don't make those three decisions, the market will make them for you,
• 13:47
  SPEAKER_00: and a buyer's agent will report the result to a committee you've never met.
• 13:55
  SPEAKER_00: Well, that's the episode.
• 13:56
  SPEAKER_00: Here's the one thing I do this week.
• 13:58
  SPEAKER_00: Pick one offer you sell today and run it through the three questions in order.
• 14:02
  SPEAKER_00: If the answers don't hold up on paper, they won't hold up with a buyer's agent either.
• 14:06
  SPEAKER_00: To download the three Ps, go to the post on Uphoff on Media on Substack
• 14:10
  SPEAKER_00: or visit uphoffadvisory.com slash frameworks.
• 14:14
  SPEAKER_00: That's U-P-H-O-F-F-advisory.com slash frameworks.
• 14:20
  SPEAKER_00: If you want to see how your own offer looks to a modeled buying committee,
• 14:24
  SPEAKER_00: that's what buying committee modeling is for.
• 14:26
  SPEAKER_00: And you can reach me through the same site.
• 14:28
  SPEAKER_00: I'll be sharing what the modeling is showing in future episodes.
• 14:31
  SPEAKER_00: And one favor, if the show is useful,
• 14:34
  SPEAKER_00: please follow it and leave a rating or review on Apple Podcasts or Spotify.
• 14:37
  SPEAKER_00: It's the easiest way for other operators to find it.
• 14:41
  SPEAKER_00: Back at it next week.