The Measurement Problem Nobody Talks About
Consumer e-commerce metrics are, in relative terms, a solved problem. You have a shopping cart. Revenue attributed to web-originated orders is the foundational number. Add conversion rate, average order value, and customer acquisition cost, and you have the beginning of a useful measurement framework. The attribution chain is direct and visible.
Most of the world's B2B commerce doesn't work this way. The website doesn't close the deal. The website starts a conversation. A customer visits, reads product specifications, decides whether this company is worth calling, and picks up the phone. The purchase happens later, off-platform, through a channel the website analytics don't capture. The website was essential to the sale and invisible in the data.
Specialty Metals Inc. (SMI) is a concentrated illustration of this problem. A small US distributor of custom-cut specialty metals — stainless steel, aluminum alloys, industrial grades — with customers who are small manufacturers needing precision materials quickly and without minimum order quantities. In 2008, SMI launched a website. They explicitly decided not to add a shopping cart. The decision was reasonable: their products required a human sales conversation before an order could be placed. You can't configure "stainless steel to .002-inch tolerance, 150mm by 400mm, same-day ship" in a standard checkout flow.
So the website existed, had traffic, and generated zero measurable revenue. Measuring it required developing a KPI framework that connected website behavior to business outcomes through the sales pipeline — not stopping at web analytics as if traffic were the goal.
What SMI Was and Why the Website Mattered
SMI's competitive differentiation was specific: expert knowledge about supplier quality, willingness to sell in small quantities that large distributors couldn't accommodate, cutting to tolerances under .001 inches, and same-day or next-day shipping for in-stock items. Their customers were small manufacturers who needed precision materials urgently, without volume commitments.
Before 2008, SMI's customer acquisition relied on direct mail, trade magazine advertising, and industrial directories. These channels had a geographic ceiling — you could only mail so far, and directory listings were discovered only by people already looking in the right place. The website lifted that ceiling. A small manufacturer in Germany could find SMI through a search query and contact them without either party having any prior awareness of the other's existence.
The four explicit goals SMI had for the website tell you what the measurement framework needed to track:
- Drive new sales — primarily through geographic reach expansion
- Make product and contact information available — reduce the pre-call friction for interested prospects
- Add legitimacy to the brand — for prospects who find SMI through search and need to trust an unfamiliar name
- Provide an outreach channel for existing clients — make it easier for current customers to reorder and to refer
None of these goals produce a direct revenue number in a web analytics dashboard. Each requires connecting the website's activity to a downstream business outcome.
The Metrics That Actually Track B2B Website Value
Request-for-quotation volume and source is the primary conversion event. SMI's site allowed prospects to submit an inquiry requesting contact. The RFQ is the equivalent of "add to cart" in a B2B model — it's the moment when anonymous interest becomes a named prospect. Tracking RFQ volume by acquisition source (organic search, direct, referral) tells SMI which channels are working and how the website is contributing to each.
But volume alone is a vanity metric. An RFQ from a prospect who wants a single 100mm stainless rod is not the same as an RFQ from a manufacturer placing recurring monthly orders. Volume metrics need to be paired with quality metrics.
RFQ-to-quote conversion rate is the quality filter. What percentage of RFQ submissions become actual sales quotes? A drop in this rate suggests the website is attracting the wrong audience — visitors who are curious but not qualified buyers, or buyers who are in a category SMI can't serve profitably. Improving the website's product copy and specification detail typically improves this rate by filtering out unqualified traffic before they submit.
Quote-to-order conversion rate tracks whether SMI's sales process is working once a prospect is qualified. This metric isn't primarily a website metric — it's a sales metric — but it connects to the website question of trust and brand legitimacy. A prospect who converted via web search and has only a domain name and a phone number to evaluate SMI against is less likely to order than one who found a professional, detailed website with clear product specifications, a real About page, and recognizable payment and shipping terms. The legitimacy goal in SMI's objective list is measurable here.
Time-to-first-contact is an operational metric that lives downstream of the website but drives website value. When a prospect submits an RFQ, how quickly does the inside sales team respond? In industrial procurement, the purchase decision often goes to the first credible supplier to respond. If SMI's website generates qualified RFQs but the response time is 48 hours, the website is generating leads that competitors are closing. A fast response rate multiplies the value of the website's lead generation.
Geographic reach measures whether the original rationale for building the website — expanding beyond direct mail geography — is actually working. Tracking where RFQ submissions originate, compared to where SMI's customers were before the website launched, reveals whether new markets are opening. An RFQ from a manufacturer in Japan is a channel that did not exist before 2008, and its presence in the data validates the website investment in terms that the original business case specified.
Return visitor behavior captures the consideration cycle. B2B purchasing decisions, particularly for custom materials, involve research and comparison across multiple sessions. A prospect who visits the product specification pages three times in two weeks before submitting an RFQ is expressing meaningful purchase intent. Tracking this behavior shows which content is doing the heavy lifting in the consideration phase and informs where to invest in the website's information architecture.
The Shopping Cart Question — When to Add Transactional Capability
The decision SMI made not to add a shopping cart was reasonable for their product at launch. Whether it remained the right decision was an empirical question that the RFQ data could answer.
The business case for a shopping cart requires a specific set of conditions. First, there must be a segment of products with stable, non-custom specifications that don't require a sales conversation to scope. Standard metal dimensions in common grades, ordered without custom cutting, might qualify. Second, there must be evidence that customers who want those standard products are reaching the website and abandoning without completing a transaction — not converting to RFQ because they know what they want and don't need a conversation. Third, the e-commerce conversion rate for that segment must be high enough to justify the product investment.
If the RFQ data shows a consistent pattern — prospects requesting identical standard specifications, no custom parameters, arriving from high purchase-intent search queries — that's a signal. A shopping cart would close those customers without a sales interaction. If the RFQ data shows that most inquiries require a sales conversation to define the specification, adding a shopping cart doesn't solve the conversion problem. It adds a checkout experience that a tiny fraction of visitors use while the majority still need to call.
The shopping cart decision is also a product scope decision that extends beyond checkout. A configuration interface for custom-cut metals isn't a standard form. It requires a product configurator that captures all relevant parameters: material grade, dimensions, tolerances, surface finish, quantity, delivery requirements. Building this correctly is a real investment in both engineering time and UX design. The business case needs to reflect the actual build cost, not the theoretical cost of "just adding a shopping cart."
What a PM Should Take From This
The SMI case teaches the discipline of building a measurement framework that matches the actual business model rather than the easiest business model to measure.
Consumer e-commerce measurement frameworks are widely documented and heavily discussed because the data is abundant and the attribution is clean. B2B measurement is harder, less discussed, and more consequential for the categories of business where it applies — which is most of the world's actual commerce by dollar value. Companies that treat their website as a brochure because they can't measure its contribution are making an error of measurement, not an error of fact. The website contribution is real; the measurement framework just isn't built for it yet.
The practical framework: map the customer's actual decision journey, not the website journey. For SMI, the journey is: search query → website discovery → product specification review → credibility assessment → RFQ submission → sales response → order. Each step in that journey is measurable. The website's contribution is measurable at each step where website behavior influences the probability of advancing to the next step. Build metrics for each connection, not just for the website session itself.
The shopping cart question is the right question to ask once the RFQ data is mature — but it's a product investment decision, not a measurement decision. Getting the measurement right first tells you whether to build it. Building it without the measurement leaves you hoping it works.