Packaging by Workflow, Not Industry: Letting Customers Self-Select

TL;DR Industry is a label on the buyer. Workflow is a description of the work. Neither is the unit you package for, because the unit is the Customer Group. The workflow axis is where Customer Groups become visible; the industry axis is where they stay hidden. When a vertical package half-fits, the failure is rarely rejection but acceptance followed by amendment, and after enough deals the amendments are the product.

A software company ships a healthcare package. Then a financial services package. Then manufacturing. The pricing page finally looks organized, and sales has a page to send after every first call. Six months later, every deal still closes with a configuration nobody published.

That gap is not a sales execution problem. It is the packaging axis reporting that it tracks nothing the customer does. This article is about the axis that works better, the mechanism it runs on, and when industry is genuinely correct. What packaging decides and which archetypes exist is treated separately.

Industry Is a Label on the Buyer, Not a Description of the Work

Grant the default what it deserves. A vertical package is legible in a way almost nothing else in pricing is: a name for a campaign, a page sales can send, a competitor it visibly beats on a shortlist. Those are real advantages, and a structure that lacks them pays for it in every deal cycle. The second driver is quieter. Industry is the axis the rest of the company already runs on, so packaging inherits it without anyone convening a meeting to choose it.

Which is the problem. Industry is a label on the buyer, not a description of the work, and packaging has to follow the work.

The label predicts the logo, not the usage

Here is the pattern we see most often once you look at usage rather than logos. Two companies carrying the same industry code run entirely different work through the same product: different capabilities, different sequences, different people, different economics. Two companies in unrelated industries run nearly identical work. The label predicted the logo on the contract and almost nothing about the usage underneath it.

The consequence is mechanical. Packaging groups capabilities into what a customer can buy, and capabilities belong together when the work travels together. An axis that ignores the work will group capabilities that have nothing to do with each other.

Be precise about what “packaging” means here, because the industry axis corrupts every version of it. Editions are one archetype. So are modules, à la carte selection, add-ons, a platform carrying apps on a shared core, a single all-in-one offer, and bundles.

Usage does not belong on that list. A usage or token limit is a metric decision, not a packaging one, and treating it as an archetype collapses the meter into the offer.

The edition ladder is by far the most common archetype, and common is not the same claim as correct. Nor does an edition ladder have to be a sophistication ladder: editions can be independent, can step up with organizational maturity, or can follow a progression that exists only in that product and that market.

Whichever archetype a company uses, the axis question is the same. Packaging needs a cluster of customers who derive value in similar ways regardless of size or vertical. That is a Customer Group, and how it differs from the labels companies reach for first, personas, buying committees, and jobs, is worked out elsewhere. When the axis is wrong, the failure surfaces somewhere specific, and it is not the pricing page.

What Happens When a Vertical Package Half-Fits Everybody

The failure people expect is rejection: the prospect says the healthcare package is not for them and walks. In the deals we have reviewed, that almost never happens. What shows up instead is acceptance followed by amendment. The buyer takes the vertical package because it carries their name, then asks for capabilities from a different package plus a carve-out on something included that they will never turn on. Sales agrees, because the deal is otherwise good. It closes at a configuration that exists nowhere in the pricebook. From there it goes one of two ways. Either the configuration gets instantiated as a new SKU, and the pricebook grows an entry nobody designed, or it does not, and the capabilities ship with no entitlement control behind them. In that second case the customer is free to use what they never paid for, and nothing in the system knows the difference.

Do that enough times and the exceptions become the product. The published structure describes what the company meant to sell; the signed contracts describe what it sells.

When the exceptions become the product

The diagnostic is available to any company with a deal record. When the same few amendments keep appearing across deals with no industry in common, they are describing a Customer Group the packaging has no container for. The sales team has been doing the packaging work by hand, deal by deal, for as long as the pattern has run. Nobody wrote it down, so nobody built it.

The cost is structural. Custom configurations are the most expensive possible way to discover a packaging boundary, because they discover it once per deal and the discovery accumulates nowhere.

It does leave a trace. The distance between what the pricebook specified and what closed deals recorded is pricebook deviation, and systematic deviation indicts the architecture rather than any seller’s discipline. Reading that gap against a real deal record is a conversation our team has with software companies constantly.

The Workflow Axis Is Where Customer Groups Become Visible

Now the correction that keeps this from becoming a slogan, because the title of this article is a half-truth on its own. The workflow axis is not a replacement unit for packaging. You do not package by workflow the way you would have packaged by industry, swapping one label for another. You package for Customer Groups. The workflow axis is the direction you look along to find them.

And the group is not the workflow either. What actually holds a Customer Group together is the value returned to the enterprise from using capabilities across workflows. The work is where you look; the value the company gets back is what you are grouping on. Follow it and what the industry axis cannot surface comes into view: which capabilities are used together and which never appear in the same account, where a customer’s own economics change as they scale, and where one group stops and the next begins. Those are the packaging boundaries. They were always there. What kept anyone from looking is a misunderstanding of willingness to pay: the belief that customers in different industries should pay differently because their willingness to pay differs by industry. That is marketing segmentation borrowed into software and almost universally misapplied, because it rides on the industry label rather than on anything about the work. Group the same accounts by the value their enterprise gets back and the picture inverts. Customers in unrelated industries often look alike, with willingness to pay close enough to treat them as one group.

Package for the group, find the group in the work

Once the groups are visible, the design decision becomes tractable. Which capabilities sit in which part of the offer and which move as buyers step up is capability allocation, and it is live inside every archetype, not only inside an edition ladder. In a modular offer the question is which module carries a capability. In a platform with apps, whether it belongs to the shared core or to one app. In an all-in-one offer, whether it is in the offer at all. The workflow view makes that allocation decidable instead of arbitrary.

It also keeps two jobs from collapsing into each other. Packaging’s commercial job is upsell: another module, a richer edition. The value metric does expansion: more of what the customer already bought. Group capabilities on an axis that ignores the work and one job ends up doing the other’s.

How the groups are derived from usage is treated in the packaging decision framework rather than here. This article is about which direction to look.

Self-Selection Is the Mechanism Packaging Runs On

Packaging is the one decision in the monetization stack a prospect meets before anyone from the company speaks to them. That sequencing is the mechanism. Self-selection means the buyer looks at the options and recognizes which one is theirs, without a salesperson explaining it. Shorter cycles, cleaner qualification, fewer amendments: every benefit attributed to good packaging assumes the buyer sorted themselves correctly first.

The test is whether the buyer knows which one is theirs

Peer-reviewed economic theory on versioning digital goods established the logic long before SaaS existed: differentiated versions let a vendor serve buyers who value the product differently without pricing each one individually. The condition attached to that result is the part that matters in a packaging discussion. The versions have to differ on an attribute buyers actually evaluate.

Peer-reviewed modeling of bundling with customer self-selection reaches the same precondition from another direction. Letting customers choose among structured options performs well for goods whose marginal cost is small relative to what buyers will pay, and it works precisely because the choosing does the sorting. That same body of work draws a distinction this argument needs: where buyers have categorically different uses for a product, differentiated versions serve everyone better, and where buyers do the same work at different intensities, the case is far more conditional.

An industry label is perceived by the buyer and does not mark a categorical difference in use. How the work runs end to end frequently does. That is why the industry axis breaks self-selection specifically. A buyer whose industry sits on one package and whose work matches another has been handed a false signal. They either pick wrong, or pick right and immediately ask for an exception. Either way the deal needs manual correction, an exception that better-structured options would not have required. If a buyer needs to be told which package fits, the packages are not differentiated on anything the buyer experiences.

None of which argues for putting everything on the page. How many options to offer is a separate question, treated in good-better-best and the choice set.

When Industry Is Genuinely the Right Axis

Sometimes the vertical package is correct. A firm that argues against it in every case, without naming the conditions where it works, is giving you its standard answer rather than a diagnosis. Here are the conditions.

Regulated work is the clearest case. When an industry’s obligations genuinely change the capability set, those capabilities travel with the vertical because the law put them there. Audit trails, data residency, retention rules, consent handling, and validated environments are not marketing differentiators presented as a package. A buyer inside that regime cannot use the product without them, and a buyer outside it will not pay for them. That is a real boundary in the work, and it happens to coincide with a vertical label.

Procurement is the second case. Where the buying process is itself organized vertically, through public sector frameworks, purchasing schedules, or consortium agreements, a vertical package can be the shape the purchase requires even when usage barely varies from the general offer. The package answers a question about how the money moves rather than how the software is used, which is a legitimate reason to exist.

The test is never whether the vertical exists. It is whether the vertical is a proxy for a difference in the work. When it is, the industry axis is not a shortcut, it is correct. How many verticals to carry and where to draw the lines is judgment specific to one company’s groups and one market, and there is no rule worth publishing.

The Questions to Ask of Your Own Packaging

None of this resolves into a procedure. It resolves into questions you can put to your own structure.

Ask what the exceptions have in common. Pull the amendments from recent closed deals and look for repeats across deals sharing nothing else. Repeats describe a package you do not sell.

Ask whether a buyer can pick without help. Put the options in front of a prospect with no seller in the room. If they do not land where you would have put them, the differentiation is invisible to the person who needs to see it.

Ask whether two customers inside the same vertical package use the product the same way. If they do not, the package is grouping by label.

Ask what the vertical is a proxy for. If the answer is a marketing page rather than a difference in the work, the package is doing marketing work.

Ask which axis the packaging actually followed. If the boundaries map to the product organization’s feature families rather than to how customers extract value, packaging followed the org chart. That is the most common architecture mistake we see, and its signature is every deal negotiated as a custom configuration.

The goal is not fewer verticals, and it is certainly not a rule about how many packages to carry. It is that whatever the packaging groups together is grouped because customers use it together. If you are working that question against your own deal record, talk to an expert and one of our pricing practitioners will reply directly.

FAQs