Marketing Case: Belden

The second case of Unit 3, again in Professor Das Narayandas's Market-Facing Functions: Belden and Digital Transformation: From Product Sales to Solutions Sales.

Olay was the story of a brand taking itself apart. Belden is the opposite — a company that has sold cable for 120 years and wants, at the very moment business is best, to put itself back together: to stop selling parts and start selling outcomes. In Capital, Marx called the conversion of a commodity into money a "salto mortale," a perilous leap. What makes Belden's leap perilous is that, at take-off, the ice under its feet is still solid.

I wrote this one before class too. I'm putting my judgements down here and will come back to check them after the session.

Here goes — enjoy:

Case Summary

November 2022. Ashish Chand, head of Belden's Industrial Automation business, is sitting on a pile of good news: the division's revenue and EBITDA have both beaten budget, and the five-year plan is ahead of schedule. But three things are weighing on him — a dispute with a French distributor, a procurement deadlock with an American metro customer, and US$26 million in budget that has to go either to sales or to R&D.

Where Belden comes from: founded in Chicago in 1902, it makes wire, cable, connectors and switches. By 2019 it had 10,000 employees, factories in 13 countries, US$2.1 billion in annual revenue, and a customer list that included BMW, Disney, ExxonMobil, General Motors, Goldman Sachs and Hilton.

Its way of growing was textbook: since 2007 it had acquired 16 companies and accumulated 15 brands. The acquired sales teams kept selling their original brands — the Hirschmann people were still selling Hirschmann; the acquired distributors kept their original distribution rights — some could sell only one brand, some several. The 15 brands had never been assembled into "one Belden."

How much IA matters: nearly half of Belden's revenue and 54% of its EBITDA. 3,735 people, 281 of them in sales. Inside were 9 product-based business units, each with its own P&L, setting its own prices, its own margins and its own strategy.

The old way of selling, in one sentence: you take a product list to meet someone holding a purchase list.

The buyers were plant maintenance supervisors, purchasing managers and project managers. They had an equipment list and a target price, and they decided based on immediate technical needs and budget. The sales process had five steps and a deal could close in a week. Pricing was cost-plus. Bonuses were based only on revenue, booked at shipment; profit wasn't part of the evaluation.

Meanwhile, IT vendors like Cisco and IBM were meeting the CIO, CTO and CEO at the same customers, talking about "what your data can do." In Chand's own words:

"IT vendors are talking to our customers' CIOs and CTOs about the possibilities of data, while we are being relegated to component suppliers."

The melting iceberg. The phrase comes from John Kotter's fable Our Iceberg Is Melting — a colony of penguins discovers that the iceberg they have lived on for generations is melting from below. Chand used it to describe IA's situation: OT hardware was commoditising and margins were falling; IT and OT were converging into "Industry 4.0," a market expected to grow from US$102 billion in 2020 to US$337 billion in 2028; value was shifting from "the wire" to "the data"; and Belden's growth in recent years had come mainly from buying companies, not from growing on its own.

In 2019 he launched the transformation, renamed the division Industrial Automation Solutions (IAS), and called the first step ESD — Enhanced Solutions Delivery.

What ESD changed:

• It picked 7 sub-industries and sized the addressable and serviceable market for each. Rail transit was the largest, with a serviceable market of US$504 million; consumer packaged goods (CPG) was piloted first.
• The 9 P&Ls were merged into 2, and in July 2022 into 1. The direct sales force was reorganised from regions and products to industries.
• Pricing authority moved from the product groups to sales. Account managers were upgraded to "solutions account managers" (SAMs), and two new roles were created: digital automation consultants (mostly former plant managers, US$230,000 a year in cost) and solutions consultants (Industry 4.0 technology experts, US$190,000). In the first phase they hired 7 of the former and 51 of the latter.
• The sales process went from 5 steps to 8, with each item checked against a MEDDICC scorecard: metrics, economic buyer, decision criteria, decision process, identified pain, champion, competition. A deal took at least 6 months, whereas a product sale took a week.
• Pricing switched to value-based. A tool called the Value Calculator took in a customer's downtime, changeover time, cleaning time and scrap rate, and calculated how much OEE improvement — and how much money — the solution would deliver. In one sample beverage plant, OEE went from 68% to 81%–88%. The average solution sold for US$200,000–250,000; the same scenario sold as products was US$150,000.
• Network audits were charged up front: US$10,000 for a single plant, US$80,000–120,000 for multiple sites. If the customer signed the full deal, the audit fee was credited; if not, they paid anyway.
• Compensation changed to 70% revenue and 30% ESD orders, credited at signing rather than at invoicing six months later.
• Two Customer Innovation Centers (CICs) were built, in Stuttgart and Santa Clara, where customers could see solution prototypes and put on VR headsets to watch how data flowed through their own plants. Chicago, Shanghai and Bangalore were in the queue.

The report card and the bill: IAS revenue grew from US$1 billion in 2019 to US$1.4 billion in 2022, with EBITDA US$84 million over budget. ESD itself: 69 customers, US$7.7 million in revenue, US$16.3 million in orders — less than 1% of the division's revenue, against a goal of half by 2026. The transformation had so far funded itself, thanks to a pandemic-era surge in small projects, maintenance and replacement orders. But the R&D bill for edge-to-cloud software was on its way, and Google Cloud had announced a manufacturing solutions unit in May 2022.

The three questions in front of Chand:

1. French distributor NTE had found a lead at an autonomous logistics company, but NTE couldn't do network audits, and it was demanding to take part and get 20% of the service fee — US$28,000. The customer wanted to deal only with Belden directly, and required delivery within 5 days of ordering.
2. A city transit authority in the US Midwest, with 8 metro lines and more than 160 stations, was willing to buy a passenger information system solution. But the solution spanned several Belden brands, and its usual distributor, Magenta, didn't have distribution rights for some of them. Making an exception for Magenta would upset other distributors; making Magenta buy through someone else would squeeze its margin. The third option was to spend US$300,000 and 6 months building a dedicated "ESD partner category."
3. US$26 million in additional budget. A salesperson costs US$250,000 a year and brings in US$2.7 million of EBITDA over 5 years; the same money in R&D brings US$2.15 million over 5 years, but US$1 million in year five alone — and the sales figure didn't account for attrition, while the R&D figure didn't account for the channel, which makes up 70% of Belden's revenue.

My Thinking

The hardest transformation starts in the year the numbers look best

When an iceberg melts, the people standing on it can't feel it — the ice under their feet is still solid.

That is exactly Belden's situation: IA contributes more than half the company's profit, revenue grew 40% in three years, and EBITDA beat budget. Not a single number is crying for help. Every danger Chand sees lies outside the financial statements: the direction of margins, who the customer's C-suite is talking to, whether growth is bought or grown.

That also dictates how he can transform. He can't "burn the boats," because that pot is cooking the whole group's profits. So ESD is a case of the old business feeding the new one: cash flow from product sales pays the new consultants' salaries; the windfall from pandemic-era replacement orders pays the CIC bills.

The speed limit of a transformation is the speed the cash cow allows.

My own read is that this is actually the more genuine kind of transformation. Most companies only move once the numbers turn ugly, and by then there's no cash cow left to feed a new business. Belden's problem is a luxurious one: when the numbers look great, how do you convince everyone that "we must change"? Chand's answer was to explain the why first — the case notes that team attrition since the transformation began has been under 7%, and he says: "Good people leave when you don't have a strategy. Once they understand why we're doing ESD, they want to stay."

Selling solutions changes not your pitch, but who you meet

The biggest change in ESD isn't "from comparing specs to asking about pain points" — that's just the pitch. The real change is that the person sitting across the table on the customer's side is someone else.

When selling products, Belden met the maintenance supervisor. He had a list, a target price, and this quarter's budget. When selling solutions, Belden has to meet the CFO, the COO, the VP of supply chain, sometimes the CEO. They don't have a list; they have KPIs — throughput, downtime, scrap rate. The budget is different too: solutions come out of capital expenditure, not the operating budget.

Fernández, who leads the solutions consultants, puts it precisely:

"We don't go to customers with a predetermined solution. So what we need are people who can ask the right questions of the right people."

There's one detail in the case I think is most worth remembering: the SAMs who close fastest are the ones who move slowest at the start. The fast-moving SAMs often discover halfway through that there's an economic buyer they've never met, and have to tear it all down and start over. The slow SAMs meet every person who has a say, and the second half goes smoothly instead.

At its core, the MEDDICC scorecard forces salespeople, before they start selling, to write down "who decides, what they care about, who will help me, who will block me." It's a discipline, not a technique.

The paid audit is where pricing power begins

Belden did something many companies "in transformation" don't dare to do: it charged for consulting up front.

Network audits cost US$10,000 to US$120,000, credited if the full deal is signed, payable regardless if not. On the surface it's a small revenue line, but in practice it does three things: it filters out customers who aren't serious; it makes the customer acknowledge "this is worth money" before they commit; and it means Belden's consultants' time is no longer free pre-sales.

The Value Calculator is the other side of the same logic. It translates the value of a solution into language the customer's CFO understands — how much OEE improves, how much is saved each year — and then pegs the price of the solution to that number rather than to the cost of the parts. In the words of sales VP Mehrotra: "When we can understand and articulate the value of a solution, we earn the right to price on value."

US$200,000–250,000 versus US$150,000: the premium doesn't come from the products getting more expensive. It comes from the conversation shifting from "how much does this cable cost" to "how much does your line lose every year to downtime."

The half that didn't change is becoming the new bottleneck

ESD changed a lot, but a few things were deliberately left alone: channel managers' responsibilities didn't change; distributors' brand-by-brand distribution rights didn't change; product KPIs didn't change; 70% of compensation still depends on revenue, and profit still isn't part of the evaluation.

The case is very honest about why. The channel accounts for 70% of revenue and handles things Belden can't do itself, like 5-day delivery: "We don't have the resources, the infrastructure, or the desire to replace our channel partners." Switch compensation to 50/50 overnight, and salespeople's income would plunge during the 6-month ramp-up, and people would leave. Piloting in CPG first and building new rules once there's enough volume is a reasonable pace, too.

But the cost has already appeared. Management noticed that SAMs generally hit their product quota first and only then work on ESD — because products count for more in their bonus. People follow the compensation plan, not the strategy speech. And not one of the three questions in front of Chand is a new problem; all of them are fruit borne by "the half that didn't change": channel conflict, brand distribution rights, revenue-driven evaluation.

It's something I also wrote about in the Olay piece, seen again from a different angle: at Olay, the strategy changed three times but the metrics didn't, so the strategy never landed; at Belden, the strategy is set but the metrics have only half changed, so the strategy has only half landed.

The three questions are really one: turning exceptions into rules

NTE, Magenta, the US$26 million — they look like three separate matters, but underneath is the same problem: how a channel system designed for selling parts can accommodate selling solutions.

On the NTE deal, I'd go with a hybrid: Belden delivers the audit and services itself, since the customer only wants to deal with Belden anyway; all products ship through NTE, which can do 5-day delivery; and NTE gets a lead referral fee, say 5%–10% of the service value, rather than 20%. Don't use the German partner — that would be telling every distributor "bring us a lead and we'll cut you out." The US$28,000 is small; the signal it sends to a channel that makes up 70% of revenue is big. Then turn this into a lead registration policy, rather than negotiating deal by deal.

On the Magenta deal, I'd do both: first make an exception for this deal, clearly labelled "verified ESD projects only" — rail transit is IAS's largest serviceable market, and you can't make a flagship customer buy piecemeal from three distributors; and at the same time immediately launch that US$300,000 ESD partner category. Solutions naturally span brands, so this kind of conflict will keep recurring. US$300,000 is roughly the size of one ESD deal, and what it buys is turning an "exception" into a "rule" every distributor can see.

Sales pays back fast, R&D pays back big, so it isn't either/or

On the US$26 million question, the table from finance is interesting: one salesperson yields US$2.7 million of EBITDA over 5 years, with US$300,000 in year one; R&D yields US$2.15 million over 5 years, almost nothing in the first two years, and US$1 million in year five alone.

But Fernández adds two footnotes that matter more than the table itself: the sales figures don't account for staff turnover; the R&D figures don't account for the channel — 70% of Belden's revenue is sold through the channel, so a new product gets amplified by that 70% channel, and a new salesperson doesn't. Stretch it to 10 years and R&D usually wins. There are also diminishing returns: not every salesperson can build a US$5 million pipeline, and once a region or an industry is saturated, each further sale gets harder.

My own inclination is a 60/40 split: 60% on the market-facing side, 40% on R&D. The market-facing money shouldn't simply hire 100 SAMs — US$26 million divided by US$250,000 is about 104 people, while the entire sales force today is only 281, ramp-up takes a year, and talent is scarce. A large share should go to consultants, who are the real bottleneck, and they should be placed in regions where there's a CIC and existing SAMs to work with. On the R&D side, invest only in edge-to-cloud software tied to use cases already validated in the CICs — the window is closing, and Google is already in.

Growth and profit will fight, and an organisation that measures only revenue can't see the fight. Solutions carry higher deal sizes, but also higher cost to serve: consultants, 6-month cycles, CICs. If you look only at revenue, you may well end up with a pile of unprofitable growth. What really needs calculating is the profit on each solution, including the additional product sales it pulls through.

One more point, not among the three questions, but which I think matters more than all three: set a date for switching compensation to 50/50. Otherwise the extra salespeople you hire will still go off and sell cable.

Takeaways for Arcus42

Reading Belden is more uncomfortable than reading Olay — because most of us recognise Belden's problems.

First, we have our own "cable" too. Every service firm has that kind of thing with a short sales cycle, that customers understand instantly, with good cash flow — an insurance policy, a structure, a single service. It's the cash cow, feeding the things that take 6 months to build: the family's asset map, the IPS, the governance structure. Belden's reminder comes in two sentences: the cash cow sets the speed limit of the transformation; but if the team's bonuses stay tied to "cable," they'll always sell cable first.

Second, charging for diagnosis is where pricing power begins. Belden's audits cost US$10,000 to US$120,000, credited against the full deal. A family office's first step — pulling together into one picture the assets, structures and risks a family has scattered across various institutions — is work of the same nature. Do it for free, and it's pre-sales; charge for it, and only then is it delivery. And for a family, what corresponds to Belden's Value Calculator is a different arithmetic: how much the same family leaks every year through duplicated risks, layer upon layer of fees, and accounts that don't know about each other. Only once you've calculated that number do you have the confidence to charge on value.

Third, families have MEDDICC too. The person who decides is never just the patriarch or matriarch. The spouse, the next generation, the CFO of the family business, the current private bank and lawyers — some are champions, some are blockers. Belden's best SAMs are the slowest at the start, meeting every single person. That's even more true for a family office: miss one person early on, and you'll have to tear it down and start over later.

Fourth, the way out of craftsmanship is a solutions factory. High-end family services today are still essentially a high-priced craft — one plan per family, each plan starting from scratch. Belden's CICs offer a very concrete answer: first turn the problems that recur within an industry into standard use cases, validate them, then customise per client. For a family office, it's about which modules — cross-border residency, trusts, insurance, governance — can be made into "standard components" first and then adjusted for each family. That is where being AI-native really ought to land.

Fifth, the best use cases come from customer conversations, not from the lab. On the US$26 million question, Belden's sales side has one very solid argument: most of the use cases in the CICs grew out of conversations between salespeople and customers, not out of R&D working behind closed doors. For a family office betting on AI, that's a reminder — however powerful the system, the founder and senior advisors must stay in front of the families. Every real conversation should flow back into the system; otherwise what the system learns is only the family as we imagine it.

Final Thoughts

Before class, the line I agree with most is "the SAMs who close fastest are the ones who move slowest at the start."

After class, I'll see which of these gets overturned, and by what.