故事 · 那个被工程师枪毙的概念 Story · The Concept Engineers Almost Killed
某消费电子团队评选三个机身概念。工程师们一致看好方案 B —— 散热结构最巧妙、装配工艺最先进,是技术上的得意之作。 可当他们把三个概念拿给真实目标客户,按客户自己在意的准则(手感、颜值、续航、价格)加权打分时, 结果让人意外:客户最买账的是被工程师认为「平平无奇」的方案 A,因为它握感最舒服、价格也亲民。 团队若按工程师偏好拍板,就会推出一台「工程上很牛、市场上遇冷」的产品。 客户选择矩阵的全部意义,就在这里:把选型的尺子,从工程师手里交还给客户。 A consumer electronics team was choosing between three chassis concepts. Engineers unanimously backed Concept B — the cleverest thermal layout, the most advanced assembly process, an engineering showpiece. But when they put the three concepts in front of real target customers and let them score against the criteria they actually cared about (feel-in-hand, looks, battery life, price), the result was a shock: the customer favourite was the one engineers had dismissed as "nothing special" — Concept A — because it had the most comfortable grip and a friendly price. Going with engineering's call would have shipped a product that "wowed the lab and flopped in the market". That is the whole point of a customer selection matrix: hand the measuring stick back to the customer.

1 加权评分矩阵:拖权重,看排名翻转 Weighted Scoring Matrix: Drag the Weights, Watch the Ranking Flip

拖左侧权重drag the weights

每个概念在每条客户准则上有个打分(1~5),乘以该准则的客户权重再求和, 就是它的加权总分。客户最在意的准则,会把分数灌给在那一项表现最好的概念。 Each concept gets a score (1–5) on every customer criterion. Multiply by the customer weight on that criterion and sum the row — that is the concept's weighted total. The criteria the customer cares about most pour their points into whichever concept performs best there.

2 加权总分排名 Weighted Total Ranking

条越长 = 客户越买账。试着把「价格亲民」权重拉到最高,或把「技术先进」拉满 —— 同样三个概念,冠军会在你眼前换人。这就是权重敏感性:选型结论对权重高度敏感,所以权重必须来自真实客户,不能拍脑袋。 Longer bar = more customer buy-in. Try pushing "affordable price" all the way up, or maxing out "technical sophistication" — the same three concepts will trade places in front of your eyes. That is weight sensitivity: the selection result is highly weight-driven, so weights must come from real customers, not gut feel.

3 现实里的客户选择矩阵 The Customer Selection Matrix in the Real World

客户视角选型:用客户的准则和权重打分,而不是工程团队的内部偏好,避免「叫好不叫座」。 Customer-perspective selection: score against the customer's criteria and weights, not the engineering team's internal preferences — avoids "critically acclaimed, commercially ignored".
加权评分:把模糊的「我觉得这个好」量化成「准则 × 权重 × 打分」,争论变成可复算的数字。 Weighted scoring: turns the vague "I just feel this one's better" into "criterion × weight × score" — arguments become numbers you can re-run.
权重敏感性:换一组权重看排名是否稳定,识别「靠运气领先」还是「真正稳赢」的概念。 Weight sensitivity: swap in a different weight set and check whether the ranking holds — separates "lucky leader" from a genuinely robust winner.
避免自嗨:逼工程团队承认 —— 决定产品命运的是客户的尺子,不是实验室里的得意技术。 Kill the echo chamber: forces engineering to admit — what decides the product's fate is the customer's ruler, not the lab's pet technology.
一句话In One Line
客户选择矩阵和 Pugh 矩阵(上一页)是一对:Pugh 帮你收敛、淘汰, 客户选择矩阵帮你用客户的价值观最终拍板。它最狠的一招,是把「权重」这件主观的事摆到台面上 —— 一旦你把权重写出来,所有人都会发现:谁说了算,全看权重给了谁。 工程师给「技术先进性」高权重,客户给「价格手感」高权重,两套权重选出两个冠军。 DFSS 的态度很明确:在概念阶段,让客户的权重赢。 但也别迷信总分 —— 它只和你填进去的打分、问出来的权重一样可靠。 矩阵是把客户偏好显性化的算盘,不是替你免去客户调研的捷径。 Customer selection matrix and Pugh matrix (previous page) are a pair: Pugh helps you converge and eliminate, the customer selection matrix helps you make the final call using the customer's value system. Its sharpest move is to put the otherwise-subjective "weights" on the table — once weights are written down, everyone sees: whoever owns the weights owns the decision. Engineers weight "technical sophistication"; customers weight "price and feel" — two weight sets, two different champions. DFSS's stance is unambiguous: at the concept stage, the customer's weights win. But don't worship the total either — it is only as trustworthy as the scores you filled in and the weights you asked for. The matrix is an abacus that makes customer preferences explicit; it is not a shortcut around real customer research.
常见误用Common Mistakes
权重由工程团队内部拍脑袋。权重必须来自真实目标客户的调研,否则就是给自嗨披一张表的外衣。 Weights set by gut feel inside the engineering team. Weights must come from real target-customer research — otherwise you've just dressed up an echo chamber in a spreadsheet.
只看总分冠军,不做权重敏感性检查。换几组合理权重看排名是否稳定,险胜的冠军要警惕。 Reading off the top total without checking weight sensitivity. Re-run with several plausible weight sets to see whether the ranking holds; a champion that wins by a hair is a warning sign.
准则太多太碎,互相重叠。选少数互相独立的关键准则,否则某个维度被变相重复计权。 Too many fragmented, overlapping criteria. Pick a small set of independent, decisive criteria — otherwise one dimension gets weighted twice through the back door.

客户选择矩阵