A good reference conversation is valuable because it belongs to someone outside the supplier's own account of the product. A prospective buyer can ask how another organisation approached the work, what proved difficult and what it would do differently. That value creates a temptation for a commercial team: when the customer is willing and the conversation is useful, ask again. A series of individually reasonable requests can become an unreasonable demand without anyone intending to impose one.
I saw that risk in our reference work at Shenqiao Systems. Different account teams were approaching some of the same customers, each with a credible reason why its discussion mattered. The customer experienced the combined demand, while each colleague saw only one request. I wanted the programme to make that total visible and give the customer a practical way to define its contribution. A reference is more credible when the person participating controls their time and the scope of what they have agreed to discuss.
Count the demand from the customer's side
In April 2026, we introduced a six-month programme with twelve consenting customer organisations and a CNY1.2 million coordination and editorial budget. Each organisation could accept up to two live reference discussions in a quarter through the programme. The ceiling was not a promise to supply two conversations. It was a limit within which the customer could still decline, defer or ask for a different format. That distinction prevented a planning number from becoming an entitlement assumed by the sales team.
The first complete quarter, July to September, produced thirty-one requests. Nineteen discussions took place, seven were deferred and five were declined. Each request had one recorded outcome at quarter end, and the nineteen completed conversations remained within the individual organisations' agreed ceilings. Those figures describe the programme's operation. They do not establish customer satisfaction, sales quality or the number of contracts caused by a reference. The wider six-month period was still incomplete when I wrote this account, so it has no final result attached.
Counting declined and deferred requests mattered. If the programme recorded only completed conversations, its report would conceal the pressure on customers and make capacity appear less constrained than it was. A refusal is useful evidence about availability and fit even when it is commercially inconvenient. I did not ask customers to provide a public explanation for declining. Their service relationship should not depend on whether they are willing to help our marketing, and an internal report should not imply that a reluctant participant is a less valued customer.
The demand account also needed preparation time. A thirty-minute call can require a customer to check what may be discussed, identify the right colleague and review the proposed subject. The visible meeting is only part of the contribution. We therefore described the likely participants and questions before seeking agreement. That allowed the customer to decide whether the request suited its experience and whether another format would be more useful. A carefully framed request can reduce wasted preparation without assuming that the supplier knows the customer's priorities.
Distinguish the permissions inside a reference request
Permission to have a conversation is not permission to publish it. A customer may be comfortable explaining a working practice to one prospective buyer while unwilling to have its comments used in a public case study. Similarly, a written account does not automatically include permission to use a company name in every campaign. I want the team to treat those as distinct decisions. Combining them into an informal idea of a reference customer makes it too easy for the original agreement to expand without another discussion.
We therefore separated live conversations, written examples and company-name use in the programme. The account owner supplied the intended subject and context, and the relevant customer contact decided what could proceed through its own approval arrangements. Marketing coordinated the request; it did not determine what the customer was authorised to disclose. If the requested subject changed, we reopened the agreement instead of assuming that a previous willingness to help covered a materially different conversation. The extra step protects the usefulness of the exchange as well as the relationship.
A reference also needs enough room to be candid. A supplier may prefer a simple account of success, but a prospective buyer often learns more from the practical difficulties and the conditions under which the work succeeded. I do not script away those qualifications to make the conversation sound more consistent with a campaign. At the same time, the customer should not be encouraged to disclose confidential information in the name of authenticity. A useful discussion has a defined subject and respects what the participants cannot reasonably share.
This is why an alternative format can be valuable without being presented as equivalent. A reviewed written example may answer an introductory question when a live discussion is unavailable. It cannot provide the same interaction or be used to imply that a particular customer has endorsed the prospective buyer's plan. The programme manager can explain those limits and offer a later conversation where appropriate. The objective is to support a considered evaluation, not to make every reference request appear fulfilled through whatever material is easiest to supply.
Keep the limit when an opportunity becomes urgent
The real test of a capacity agreement arrives when an important opportunity needs an immediate conversation. An account leader may argue that this request is different, or that the customer would probably agree if approached directly. I understand the commercial pressure, but a limit that disappears at the first urgent request is not a limit the customer can rely on. Exceptions need the customer's explicit willingness and a fresh view of the burden, rather than an internal decision that the potential deal justifies ignoring the arrangement.
I also want priorities discussed before urgency becomes the only criterion. Different opportunities may need different forms of evidence, and the most valuable conversation is not always the one with the largest potential contract attached. The programme can consider relevance, timing and the question the buyer actually needs answered. It should not promise access to a customer as though that access were controlled by the supplier. Commercial judgement includes recognising when a request cannot be met responsibly and explaining that position without blaming the customer who set the boundary.
For the CMO, the programme creates a different kind of accountability from collecting endorsements. I am responsible for the quality of the arrangement we offer, the clarity of our requests and the accuracy with which we report what happened. I cannot claim ownership of another organisation's professional voice. The nineteen conversations in the first quarter are useful activity, but the seven deferrals and five refusals are equally part of the record. Preserving that whole account is how we keep a reference relationship credible when commercial demand would otherwise encourage us to treat someone else's time as available stock.