The Shared Commercial Engine

The handoff point

Four levels of overlap appear across the cases, and they predict how much of the funnel stays central.

The useful part of this is the top two rows, because that is where most portfolios actually sit and where they consistently overestimate themselves.

Same market is weak overlap

Two businesses in adjacent categories share reputation and category knowledge. They do not share a buyer, and nothing commercial transfers beyond method.

Same customer company is often still weak overlap

This is the row that fools people, and two of the largest cross-sell theses in American corporate history sat in it.

Cendant was built on it. The 1997 merger proxy told shareholders the combination of HFS brands reaching more than 100 million customers with CUC's direct marketing expertise and roughly 69 million members would produce new growth through direct marketing of memberships into the hotel base. The same document disclosed the risk that anticipated cross-marketing benefits and technological efficiencies would not be fully recognized.Fact A detail from the SEC record is independent of the accounting fraud that followed and speaks directly to the point: after closing, the company attempted to conduct its business with parallel but separate financial reporting structures for the two legacy halves.Fact The organizations that were going to cross-market to each other did not merge their books.

Citigroup ran the same test at larger scale. The 1998 Travelers merger was justified publicly on selling insurance and banking to a shared retail customer. Citigroup spun off Travelers Property Casualty in 2002 and sold Travelers Life to MetLife in 2005.Fact · cause inferred

Same named buyer is where it starts to work

McKinsey's survey work on post-merger cross-selling contains the cleanest quantification available. In one technology services organization, accounts where salespeople had strong relationships with relevant decision-makers reached an 80 percent cross-selling rate within a year, while accounts requiring reps to sell to unfamiliar people took roughly 18 months longer to achieve similar results.Fact Same company, same products, same salesforce. The only difference was whether the seller already knew the person.

Same buyer plus same purchasing motion is where permanent central selling becomes viable

The accounting firms selling consulting into audit relationships through the 1990s worked well enough to become a regulatory problem, and Sarbanes-Oxley substantially prohibited it in 2002.Fact The second product went to the same individual, in the same building, during an engagement already underway, from someone that individual was already paying.

That is the bar. Not the same industry. Not the same company. The same person, mid-relationship, buying something similar the same way.

One handoff per company

The clearest case of a temporal boundary is eFounders, the SaaS studio now operating as Hexa, because it published its own operating model over a decade and revised it in public.

The first letter, from September 2013, describes an ambition to pool more across the portfolio and notes in passing that they rarely made cross-selling campaigns, with hosting, technology, and vendors all chosen independently by each startup at the time.Fact The founder allowed that not everything they tried would work.

Seven years later the model had not expanded. It had settled. Letter #8 reports that the way of working remained relatively untouched in nine years: the core team works with founders for 12 to 18 months with the goal of making the company independent from the studio, assembling around ten people, building a first version, and reaching small traction, after which the company raises a seed round and leaves.Fact TechCrunch described the same architecture from outside in 2023, noting that the studio supplies product design, go-to-market strategy, hiring, and fundraising for roughly the first year, then moves on.Fact

More than forty companies came through it. By the 2022 restructuring, portfolio companies had hired 3,000 people and reached a total valuation of $5 billion.Fact

What matters is the shape rather than the outcome. The studio tested cross-selling early, found little, and stopped. What it kept was commercial work during formation, on a fixed clock, followed by complete transfer. Nine years without structural change is unusual stability for anything in venture.

The handoff was not the failure of the shared model. The handoff was the model.

One handoff per transaction

The temporal version is not the only one, and it may not be the more common one.

Private equity home services platforms are the most controlled test available of whether similarity eliminates the boundary. Everything is held constant. Common ownership. Nearly identical businesses. The same buyer, which is a homeowner with a broken system. The same purchasing motion. A stated strategy of standardization.

Practitioner descriptions of what these platforms centralize are consistent across independent sources: dispatch systems, call-center discipline, centralized procurement, technician recruiting, management reporting, pricing governance, along with brand and digital acquisition.Fact The consolidation thesis is explicitly to transform numerous small independent businesses into unified platforms with centralized purchasing, marketing, and operational efficiency.Fact

Then look at where revenue is created. The same sources describe residential growth as depending on call-center conversion, local marketing, financing availability, replacement demand, and technician productivity.Fact The center generates the lead and books the appointment. The technician sells the replacement job standing in the customer's basement.

At maximum similarity, with everything else centralized, the close stayed local.

Franchise systems show the same architecture with no ownership involved at all. The franchisor permanently owns brand, the national advertising fund, lead generation systems, customer database tools, pricing guidance, training, and methodology. Marriott's filings describe the mechanics plainly: franchisees contribute to the national marketing and advertising programs, and pay fees for use of centralized reservation systems.Fact Contribution rates are published and specific, for example 0.7% of weekly gross sales toward local digital marketing at one home services brand.Fact

The franchise industry has also priced this finding, which is the strongest independent corroboration in the research. From practitioner guidance on how marketing royalties are set:

Business-to-business and service brands are often a more considered purchase with a relationship sell. Marketing royalties for these brands are considerably less, usually 1-3% of sales. Why? Because they rely on the franchisee, or their sales team, to make face-to-face contact with the decision-maker.

Franchising.com practitioner guidance on marketing fund structure

An entire industry converged on charging less for the central commercial system where the sale requires a relationship, wrote the reason down, and embedded it in standard contracts. Nobody in this industry was participating in a debate about portfolio strategy. They were setting a fee.

Maximum similarity pushes the handoff farther down the funnel. It does not necessarily remove it.

The exception that is not an exception

A reader who has worked in a large portfolio will object that some organizations genuinely do run one commercial team across many businesses. They do. Three of them are worth inspecting because the inspection is where the mechanism becomes visible.

Match Group runs central advertising sales across brands that compete with each other for the same users. Match Media Group offers advertisers with budgets of $25K+ a single point of contact to run campaigns across the portfolio or on specific sites.Fact The rest of the structure is common scale in payments, analytics, moderation, experimentation, and certain back-office functions behind brand autonomy at the front.Fact The central salesforce works because the advertiser does not care which app supplied the impression.

Marriott is the more important case, because it removes the variable everyone assumes is doing the work. Sales Force One, launched in 2008, is a companywide structure built around sharing sales resources, clearly defining roles and responsibilities, tying compensation to results, and pursuing under-penetrated high-value accounts.Fact It sells across more than thirty brands and thousands of properties Marriott does not own.

Ownership is not the variable.

A franchisor with no equity in the local business can run permanent central sales, and a holding company with full equity in every unit routinely cannot. What makes the central function possible is common demand architecture, meaning a buyer purchasing a standardized unit the same way each time. Corporate travel managers and meeting planners buy room nights. Which brand supplies them is a detail.

Dotdash Meredith contributes the sharpest structural observation. One commercial organization sells across forty-plus brands as unrelated as People, Investopedia, Allrecipes, and Byrdie. After the merger, the combined company adopted an organization where sales teams are organized around content verticals rather than around publishing brands, with flexibility where some advertisers do not fit into one specific content category.Fact

The salesforce did not organize by business. It organized by advertiser.

The unit of commercial organization follows the buyer, not the business.

What happens when you push past the boundary

WPP ran the experiment at maximum scale and left a public record.

Sorrell's horizontality produced bespoke cross-agency client teams. Team Detroit, five agencies working together for Ford from 2006, became Team GTB in more than 40 offices with more than twenty WPP companies on the account by 2017, a structure its own leader described on stage as complicated, noting it included no fewer than five PR agencies.Fact

The first half worked. An industry assessment in 2013 concluded that as a new business tool the structure had indeed won and protected several assignments, while evidence on delivered integration was much more mixed and many high-profile team structures had fallen.Fact Contemporary analysis named the delivery cost: the dedicated agency approach forgoes access to best-of-class talent.Fact A client who buys the holding company gets whoever the holding company assigns.

Two things closed it. Integration had limits because the agencies often had their own P&Ls, and there were more than 160 of them.Fact And the structures depended on named individuals. WPP built Enfatico globally to serve Dell; the clients who agreed the deal left Dell, and Enfatico expired and was folded into another agency.Fact

By 2018 the term was retired, with Mark Read saying WPP was ditching the word because it wasn't well understood.Fact Within a week of Sorrell's departure, Ford called a creative review.

WPP could centralize access and the right to compete for a global account. That is a real asset. It could not centralize the specialized work required to serve the account, and it could not solve who got paid for what.

The opposite extreme is worth naming briefly. Constellation Software has acquired hundreds of vertical software businesses, operates a head office of roughly fourteen people, and does not centralize sales, product development, or human resources.Fact It shares capital allocation, acquisition process, and benchmarks. It is among the best long-run compounders of the period, and it centralizes nothing customer-facing at all.

What the center is actually good at

Across every structure examined, the same division appears.

The line between the second band and the third is domain credibility. A person covering seven businesses is credible in none of them at the level a buyer spending real money requires.

The method for building a sales playbook can be shared. The sales playbook itself usually encodes a specific buyer.

Why these programs fail

Four failure modes recur, and they are economically meaningful rather than merely untidy.

That last finding is the one worth sitting with. A half-committed cross-portfolio commercial program consumes seller attention and creates territorial conflict without producing revenue. The choice is between doing it properly and not doing it.

What compounds

Some things transferred reliably across portfolios of every kind studied.

Method. Benchmarks. Recruiting capability. Operating knowledge. Institutional standing. Shared procurement. Demand generation infrastructure where the buyer is common. Experience establishing a first commercial motion.

Some things generally did not transfer across divergent buyers. Customer relationships. Active pipeline. Mature quota ownership. Product-specific closing expertise.

Distribution itself is less transferable than the capability to repeatedly create it.

That is a tendency rather than a law. Match, Marriott, and media networks are real exceptions where shared demand architecture lets considerably more transfer, and any portfolio sitting at that level of overlap should behave differently from the ones that are not.