What I changed in the material, the measure and the request

In January 2024, a proposal to increase general market exposure was being discussed alongside a list of opportunities that had not advanced. Peiyu Qiao wanted more support for account teams, and some marketing colleagues favoured a broad campaign that could be launched quickly. I was concerned that we were treating an absence of movement as an absence of awareness. Before committing the larger programme, I authorised CNY1.8 million for a twelve-week review of thirty-two stalled opportunities selected with sales.

The review combined account records with interviews that the participating customer contacts agreed to provide. We recorded the main unresolved issue as understood at the review date, rather than claim to discover a single permanent cause. Seventeen cases involved different operational priorities within the buying group, nine involved budget timing and six lacked an agreed executive sponsor. The categories totalled thirty-two because each case received one primary classification; secondary issues remained in the notes. The selected sample was neither random nor representative of every prospect.

Those findings changed my recommendation. Instead of using the review as a new segmentation slogan, we prepared discussion material that separated the operating problem, the proposed change, the assumptions behind its value and the decisions still needed internally. Sales retained responsibility for the account conversation. I also asked for a route through which an account team could report that our classification was wrong or had become outdated. A research label should not prevent a customer from describing a changed situation.

The concession was time and a less immediately visible campaign. The broad programme did not begin during those twelve weeks, and the review required customer and account-team attention. I do not attach a subsequent conversion rate to this case because we had no comparison that would isolate the effect of the material. The recorded result is a changed marketing recommendation grounded in a bounded review. It taught me to ask what a buying group still needs to decide before assuming that another reason to notice the supplier is the missing intervention.

In September 2024, an engineer considering our software needed to complete a contact form to obtain several introductory technical explanations. Marketing could count the submissions, but the count did not show whether the exchange helped a person assess relevance. I proposed opening selected material. The demand-generation manager objected that removing forms would make an established channel appear smaller, while technical colleagues were concerned that public documents could be used without their necessary qualifications.

We reviewed twelve documents individually with product, security and commercial owners. Six were suitable as public explanations after revision; six remained controlled annexes for a more specific exchange. The ten-week programme cost CNY1.4 million, covering editorial work, specialist review and publication. Each public document carried its scope, owner and revision date. Opening an explanation did not make confidential configuration, security or contractual information public, and the controlled annexes were not presented as material anyone could obtain merely by supplying an email address.

I agreed before release that the six public documents would no longer carry a form-submission objective. Over the first ninety days, the site recorded 4,800 document views and forty-one separately identified enquiries that cited one of the explanations. Views were events, not necessarily unique people; the same reader could return. The enquiries were not attributed sales, and we did not infer that anonymous readers became buyers. This account gave the team two different observations rather than combine them into a single inflated lead total.

The decision gave up contact records and made content maintenance more visible. A public document could circulate after someone saved a copy, so revision information and a route to the current version mattered. I kept the document owners involved after launch instead of treating publication as the end of their responsibility. The result was a more open way to begin a technical evaluation, with a smaller claim about what marketing knew of its audience. I would revisit the arrangement if specialist review found that a document's scope was no longer appropriate; openness was a considered choice for six explanations, not a rule that every technical detail should be published.

Our proposed planning calculator began with standard values intended to make a workshop move quickly. By February 2025, I was uncomfortable with how easily the resulting number could be repeated as if it described the customer's operation. A commercial sponsor preferred the consistent output because it gave account teams a clear starting point. Finance and delivery colleagues pointed out that downtime, staffing and the cost of changing a work process varied materially between customers.

I authorised a CNY960,000, eight-week redesign with those colleagues. We replaced the pre-filled conclusion with an editable workbook that separated measured inputs, customer estimates and unconfirmed assumptions. It retained a worked fictional example for learning, clearly apart from the customer's own scenario. A blank input stayed blank; the workbook did not silently substitute the example value to produce an apparently complete result. The customer could also record a cost that reduced the estimated benefit rather than choose only inputs that supported the purchase.

Eight buying groups agreed to review the prototype. Five provided enough inputs to complete a first scenario; three left it incomplete because they could not yet support important assumptions. We recorded the missing inputs and the questions participants raised. We did not describe five completed worksheets as five approved investments, measured savings or evidence that the software would deliver the modelled result. No customer data or scenario was reused publicly without a separate permission process.

The new workbook was less convenient for a short presentation. Account teams had to explain why an incomplete calculation could be a useful result, and some discussions took longer because a previously hidden assumption became a real question. I accepted that cost. Marketing's contribution was a method the customer could inspect and challenge, not ownership of the customer's investment decision. The case changed how I review commercial tools: I now ask what the document does when a necessary fact is unknown. If it still produces a confident answer without showing that gap, its apparent simplicity may be concealing the work the buyer needs to do.

In August 2025, a proposed identity refresh would have replaced four established product-family names with a single corporate naming pattern. The creative proposal was consistent and easier to present. Product colleagues were less certain that the change would help customers who used the existing names in training, support requests and internal procurement records. I asked that we examine those uses before allowing visual consistency to decide the whole programme.

The CNY2.7 million, fourteen-week project included sixteen customer interviews and review of twenty-four support and training artefacts selected by the relevant teams. These were qualitative inputs, not a representative survey. Eleven interviewed contacts described an active internal use of at least one existing product-family name; five did not identify such a dependency in their setting. The artefact review showed where a change would require coordinated revision rather than a new cover alone. We did not treat the interview count as a market preference score.

I recommended retaining the four product-family names and adding the corporate identity through a consistent endorsement and explanatory hierarchy. The company could present a coherent group while customers kept recognisable terms for the software they used. The creative team disagreed with the reduced purity of the naming system, and I accepted a more complex design specification. Product and support owners agreed the terminology; marketing did not use the project to rename technical functionality or imply a change in contractual entitlements.

The revised identity was released in December 2025 with a mapping guide for commercial and service teams. That publication date records an output, not an independently measured improvement in recognition or support performance. We kept a list of material that still needed routine replacement rather than claim that every old document disappeared at launch. The lesson was that a familiar name can be part of a customer's working knowledge. A corporate narrative has to accommodate that knowledge if it is to make the relationship clearer, even when the resulting system is less elegant on a single presentation slide.

By April 2026, several of our most willing reference customers were being approached by different account teams for similar conversations. Each request could be reasonable in isolation while the combined demand became difficult for the customer to manage. Sales wanted quick access when a prospective buyer asked to hear from a peer. I proposed that we make the total demand visible and agree capacity before treating availability as an informal favour that could be renewed indefinitely.

We introduced a six-month reference 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 under the programme; this was a ceiling, not an obligation. Written examples, logo permission and a live conversation required separate agreement. Account owners supplied the intended subject and participants before the customer decided. Declining a request did not affect the service relationship or create an expectation that an employee should explain the refusal publicly.

During the first complete quarter, July to September, the programme received thirty-one requests. Nineteen discussions took place, seven were deferred and five were declined. Those outcomes total thirty-one; no request was counted in more than one category at quarter end. The nineteen completed discussions stayed within each customer's agreed ceiling. We recorded the limits without inventing satisfaction scores or attributing contract wins to participation. The wider six-month programme was still running in October, so this profile does not give it a final outcome.

The arrangement reduced the account team's freedom to seek an immediate introduction. Some colleagues felt that a scarce reference should be reserved for their most urgent opportunity; I wanted that priority considered alongside the customer's consent and existing commitments. The programme manager could propose an alternative written account or a later conversation, but could not override a refusal. My conclusion so far is about the quality of the arrangement, not a claim of commercial return. A reference remains credible when the customer controls its contribution and the supplier remembers that someone else's professional time is not its marketing inventory.