Build a workstream library and you immediately face a question nobody warns you about: what are the top-level categories? Marketing and sales? Demand generation, revenue operations and enablement? The customer lifecycle?
We went looking for the standard answer. There is not one, and the disagreement between the major published models is structural rather than cosmetic.
A shared workload, not a taxonomy
Put the published models side by side and the pattern is clear. Some firms organise around functions, some around commercial capabilities, some around the customer lifecycle, and the academic literature around the interfaces between functions.
What the published models expose
- McKinsey packages commercial capability around customer experience, sales, marketing and pricing, with customer lifecycle management separated in the broader practice.
- Bain exposes route to market, pricing, account management, marketing, sales management and commercial operations as distinct services.
- BCG defines commercial excellence principally across marketing, sales and pricing.
- Gartner and Forrester place many of the same activities inside a cross-functional revenue operations model, with Forrester explicitly adding partner ecosystems and customer success.
Note what moves between them. Pricing is a top-level discipline for three of those firms and invisible in some others. Customer success is a named category for Forrester and folded elsewhere. Partner and channel appears and disappears entirely.
Where the boundaries are contested
Demand generation is relatively stable as a concept until a lead becomes a revenue object. The argument starts at qualification, handoff, pipeline ownership and measurement, not at whether demand generation exists.
That is the pattern across every contested boundary we found. Nobody disputes that the work happens. They dispute who owns it, when it transfers, and which function's number it lands in. BCG's revenue operations analysis identifies inconsistent data, duplicated metrics and unclear ownership as the recurring symptoms, which are handoff problems rather than taxonomy problems.
Outcomes follow interfaces
The strongest available evidence does not show that one taxonomy beats another. It shows something more useful.
Homburg, Jensen and Krohmer examined 337 European companies and identified five marketing-sales interface archetypes. Their strongest-performing configurations were characterised by strong structural linkages between marketing and sales, and high market knowledge within marketing. Le Meunier-FitzHugh and Piercy separately found a direct positive relationship between collaboration across that interface and business performance.
The lesson is about the design of the interface, not the name over the door.
The practical consequence for a workstream library is that the categories matter less than whether each workstream names its dependencies on the others. A scope item that says who hands what to whom, and when, is doing the work the research says predicts outcomes.
What your own model omits
The exercise worth running is not choosing the correct taxonomy. It is putting your library next to three published models and noting what has no home in yours.
A marketing-led structure with a revenue-intelligence front end and a sales-alignment back end is coherent. It also tends to leave pricing, partner and channel, and customer success without dedicated top-level disciplines. That is a defensible scope choice for a firm that does not sell those services. It becomes a problem only when a client's actual constraint sits in the gap and nothing in your library names it.
Choosing your own top level
If no published taxonomy is correct, the decision reverts to you, and it is worth making deliberately rather than inheriting it from whichever deck you saw last.
Three constraints are worth weighing. The first is what you actually sell. A top-level discipline with nothing beneath it advertises a gap. If you do not do pricing work, a pricing discipline containing one thin workstream is worse than no pricing discipline at all, because a buyer reads the empty shelf rather than the full ones.
The second is how your buyers describe their own problem. Categories exist to help a reader locate themselves. A structure organised around your delivery teams rather than the client's language forces every buyer to translate before they can tell whether you address their issue.
The third is where your engagements actually break. If handoffs between demand generation and sales are the recurring failure in your delivery history, that seam deserves to be visible in the structure rather than buried inside two adjacent categories that never reference each other.
What would make this wrong
The strongest counter-argument is that operating model choice clearly can matter, and a taxonomy is part of that choice. It would be too convenient to conclude that structure is irrelevant just because no single structure wins.
Homburg and colleagues do not find all commercial configurations equivalent. The better-performing archetypes have stronger structural linkages between marketing and sales, which means the way you draw boundaries can help or hinder the very interfaces that predict performance. Gartner's customer success research finds that a poorly defined charter creates duplicated effort and customer confusion, which is a taxonomy problem expressed as an operational one. BCG argues that integrated revenue operations can standardise reporting, metrics, tools and end-to-end processes.
Read together, those say the labels are not arbitrary even though no label set is canonical. A structure that obscures a handoff is worse than one that exposes it. What the evidence does not support is the claim that adopting a particular firm's category list will improve your outcomes, which is usually what a published taxonomy is being sold as.
Common questions
Should we just adopt a published model?
Only if you sell the whole of it. Adopting Bain's category list while offering three of its services produces a library with permanent empty shelves, which reads worse in a proposal than a narrower structure that is fully populated.
Does the taxonomy affect delivery outcomes?
Not that we could find. What affects outcomes in the available evidence is the strength of the linkage between functions, which your taxonomy can support or obstruct but does not determine on its own.
How do we handle a client whose internal model differs from ours?
Map to theirs in the document and keep yours underneath. The reporting friction created by asking a client to translate your categories into their own is real, and it lands on them every month rather than once.
How many top-level disciplines is right?
Few enough that a buyer can hold them in their head, and no fewer than the number of genuinely different sales conversations you have. The published models range from three to seven, which is a reasonable band to sit inside.