Skip to content

What does research say about B2B e-commerce?

There are many figures on B2B e-commerce. Many come from vendors and cannot be traced to a study. Here are the figures that hold up and what they show.

Summary

  • Buyers want to buy on their own. 67% of B2B buyers prefer to buy without a sales rep.4
  • E-commerce is the largest channel. Among B2B sellers with e-commerce, it accounts for 34% of revenue.1
  • Friction costs customers. More than half of B2B buyers are likely to switch suppliers if buying is not smooth.1
  • Easy buying means bigger buying. Buyers who get help to buy are three times as likely to choose the bigger option.2
  • Product data must be consistent. 69% of buyers see differences between the website and what the sales rep says.3
  • Search often falls short. 56% of reviewed B2B sites only return results for exact words.5
  • Self-service only saves money when it solves the problem. Only 14% of issues are fully resolved without help.8
  • What is missing. No study measures what better B2B e-commerce is worth in money. That evidence comes from your own before-and-after measurements.

Buyers want to buy on their own

  • Without a rep. The share of B2B buyers who prefer to buy without a sales rep rose from 61% to 67% in one year.34
  • But not for everything. Buyers prefer self-service when they look for information. When they judge whether a product fits, they still want a sales rep.3
  • A third each. At every stage of buying, one third want to meet in person, one third want remote contact and one third prefer digital self-service.1
  • Large orders too. Buyers' comfort with placing large orders remotely or online has grown, including orders above $500,000.1

E-commerce does not replace the sales rep. It takes over what the buyer would rather do alone: find, compare and reorder.

E-commerce is the largest channel

McKinsey's B2B Pulse 2024 is based on answers from nearly 4,000 decision makers in 34 sectors and 13 countries.1

  • Most offer it. 71% of B2B companies offer some form of e-commerce.
  • Largest source of revenue. Among them, e-commerce accounts for 34% of revenue, more than in-person sales.
  • Most effective. For the fourth year in a row, those who sell online rank e-commerce as their most effective channel.

Friction costs customers

  • Switching suppliers. More than half of buyers say they are likely to switch suppliers if the experience is not smooth across channels.1
  • Digital is a common reason. 54% of those likely to switch cite poor digital experiences.1

The figures describe what buyers say they would do, not measured switching.

Easy buying means bigger buying

  • Three times as often. Buyers who receive helpful information that eases the purchase are three times as likely to buy the bigger, more expensive option, with less regret.2
  • Confident buyers make better deals. Buyers who are confident in their decision are twice as likely to report a high-quality deal.4

Gartner does not state the sample for the 2019 figure.

Product data must be consistent

  • Mixed messages. 69% of buyers see differences between what the supplier's website says and what the sales rep says.3
  • Wrong language. On more than 65% of the reviewed multilingual B2B sites, items dropped out of search because of the wrong language in the product data.5
  • AI reads it too. 45% of B2B buyers used AI during a recent purchase.4 An AI assistant can only answer correctly if the product data is correct and readable.

Product data is the foundation for search, filters, product pages, configurators and AI. Errors in the data show up in every channel at once.

Search often falls short

  • No hit means not stocked. When a model-number search returned nothing, users quickly concluded the site did not carry the product.6
  • Search comes first. Roughly half of users prefer search for finding products.6
  • Exact words only. 56% of 23 reviewed B2B sites only returned results when the search term exactly matched the product text.5
  • Dead end. Almost 80% of the B2B sites offered no way forward when a search had no results.5
  • B2C is ahead. The consumer sites scored 1.31 times higher than the B2B sites.5

Baymard has tested more than 170 sites, mostly consumer sites. VML's sample is small, the page is undated and VML sells e-commerce implementation. The picture is still the same.

Self-service only saves money when it solves the problem

  • Cost per contact. A customer service contact costs $8.01 on average, versus about $0.10 for self-service.7
  • Few issues are fully resolved. Only 14% of issues are fully resolved in self-service, and only 36% of the very simplest.8
  • The customer cannot find the answer. The most common reason self-service fails is that the customer cannot find relevant content, in 43% of cases.8

There is no published figure for the share of customer service issues that e-commerce could solve. That figure is in your own ticket system.

AI that makes things up is a risk

  • The company is liable. Air Canada's chatbot gave a customer wrong information about a discount. A Canadian tribunal found that the airline had not taken reasonable care to make the bot accurate, and ordered it to pay.9
  • Retrieved facts reduce errors. In a study from Facebook AI Research, the share of made-up answers fell from 68% to 8% when the model retrieved facts before answering.10

An AI assistant that retrieves answers from your own systems answers with items and prices that exist. That lowers the risk but does not remove it.

What this means in practice

  • Measure before and after. Conversion, order value, repeat purchases and the share of searches with no results. That is the only evidence that applies to your business.
  • Count tickets. How many customer service tickets are about finding an item, a price or an order status?
  • Start with product data. It affects search, product pages, configurators and AI at once.
  • Keep the sales rep. Let e-commerce take what the buyer would rather do alone, and the sales rep what requires judgement.

An example of what this can look like in practice is Find the right item without the part number.

Next step

Want to discuss B2B or something else? In half an hour we go through what you want to solve and how it can be done.

Book an intro call

  1. McKinsey & Company, "Five fundamental truths: How B2B winners keep growing", September 2024. B2B Pulse Survey of nearly 4,000 decision makers. ↩↩↩↩↩↩↩

  2. Gartner, "Gartner Says Marketers Must Focus on Helping Customers in Order to Remain Competitive Today", press release, April 2019. ↩↩

  3. Gartner, "Gartner Sales Survey Finds 61% of B2B Buyers Prefer a Rep-Free Buying Experience", press release, June 2025. 632 B2B buyers, August–September 2024. ↩↩↩↩

  4. Gartner, "Gartner Sales Survey Finds 67% of B2B Buyers Prefer a Rep-Free Experience", press release, March 2026. 646 B2B buyers, August–September 2025. ↩↩↩↩

  5. VML, "The State of B2B Site Search". 23 B2B and 11 B2C sites against 22 criteria. The page is undated. ↩↩↩↩↩

  6. Baymard Institute, "E-Commerce Search Query Types", updated 2026. More than 170 sites and apps. ↩↩

  7. destinationCRM, "Gartner Survey Finds Self-Service Insufficient", November 2019. Gartner's survey of customer service leaders. ↩

  8. Gartner, "Gartner Survey Finds Only 14% of Customer Service Issues Are Fully Resolved in Self-Service", press release, August 2024. 5,728 customers, December 2023. ↩↩↩

  9. McCarthy Tétrault, "Moffatt v. Air Canada: A Misrepresentation by an AI Chatbot", February 2024. On the decision 2024 BCCRT 149. ↩

  10. Shuster et al., "Retrieval Augmentation Reduces Hallucination in Conversation", Findings of EMNLP, 2021. ↩