Client concentration & who sets the price

The survey asked whether any client bought more than 50% of the respondent’s product or service sales monthly. We divided complete answers into those reporting such a client & those reporting none. Within each group, we then examined who took decisions about product or service prices: the respondent, a business partner, a supplier, a client or someone else.^[2]

In the group reporting the monthly concentration, 67.9% named themselves, 0.9% a business partner, 4.3% a supplier, 21.7% a client & 5.3% someone else. In the other group, the corresponding shares were 84.4%, 1.0%, 3.2%, 4.5% & 6.8%. The table keeps every category visible because a story about clients alone would leave out how most respondents described their situation.

The majority answer does not mean every owner could name any price & have it accepted. The question identified who took the price decision; it did not measure how much room that person had to negotiate, whether the price covered costs or whether a proposal succeeded. Its useful contribution is more specific: a concentrated customer mix & a respondent-led price decision appeared together in many answers.

There were 943 complete answers in the concentrated group & 5,809 in the other group. The percentages apply the survey’s final weights, so they are not simply the shares of those sample counts. Rounded values total 100.1% & 99.9%, respectively; the unrounded values total 100% in each group.^[3]

The client category shows a clear difference between the 2 groups: 21.7% compared with 4.5%. Subtracting the latter from the former gives 17.2 percentage points. This describes how the reported price decision varied alongside the monthly customer-concentration answer. It does not establish that concentration caused a client to decide prices.

The 95% confidence interval for that gap was 13.4–21.0 percentage points. We calculated the difference’s uncertainty directly, using the survey’s weights & sampling design.^[3][4]

The table below shows a separate interval around the client share in each group: 18.3–25.6% where the monthly concentration was reported, & 3.8–5.4% where it was not. These are pointwise 95% intervals, meaning each concerns its own estimate. The interval for the 17.2-point difference was calculated directly, rather than inferred from whether these 2 intervals overlap. The download includes intervals for all 10 actor shares.

One distinction belongs beside this result. The concentration question referred to a client buying more than half of monthly sales. The price question offered the general answer “a client.” It did not identify that client or connect the 2 identities. We can report a greater share of client-led price decisions in the concentrated group; we cannot assign those decisions to its largest buyer.

Consider 2 invented service businesses. In each example, one customer accounts for 60% of sales in the month being reviewed. The businesses, the share & the arrangements below are fictional illustrations of a decision, not respondents or results from the survey.

The first is a product-photography studio. Its owner sets package prices & sends clients a proposal defining the number of products, images, revisions & delivery dates. The large customer chooses whether to accept. When preparing a price change, the owner can start with the package: which work is included, which additions are priced separately & when the proposed terms would apply. The customer’s share of sales remains relevant to the portfolio review, but it does not supply the contents of the new proposal.

The second is a self-employed service provider whose large customer specifies the price for each assignment. The provider decides whether to accept the work. Here, preparing a proposal starts with understanding which parts of the arrangement are open for discussion. A revised scope, a different delivery schedule or a separately priced addition might be topics to raise. These are possible conversations, not outcomes the survey demonstrated or changes it showed a customer would accept.

Both examples begin with the same monthly share. They end with different preparation because the price-decision process differs. A list sorted by customer sales would not reveal that difference. Nor would labeling both businesses “concentrated” tell the person writing a proposal what to write.

The reverse combination deserves attention too. In the survey group reporting no client above half of monthly sales, 4.5% still named a client as the price decision-maker. An owner cannot settle the price question simply by checking that no customer crosses that line. The reported relationship makes the concentration question informative, while the answers within both groups show why it is incomplete.

For a creator selling services to businesses, the examples offer questions to bring to a real brief: who proposes the amount, who decides it & what work does it cover? They do not establish how common these arrangements are among creators. The measured estimates belong to the broader Chilean survey population. Apply the worksheet to your own records & conversations rather than treating either percentage as a forecast for your client list.

A practical response is to keep 2 pieces of information together: the customer’s place in the sales mix & the way the price is decided. This is our suggested use of the finding, not a process tested by the survey. Its purpose is to make the proposed conversation specific to the relationship you actually have.

Start the customer review with a stated month & a consistent sales measure. Use the same measure for the customer & the business total, record the source & note any missing amounts. The survey’s monthly question did not provide an exact sales share for each respondent. Your own records may let you calculate one, but a calculation for a selected month should retain that month rather than become an unsupported claim about the whole year.

For example, a creator might count assignments when planning capacity, while this survey question concerned sales. Several small assignments & one high-value assignment can tell different stories about workload & sales. If you keep both measures in your records, label each clearly. Keep percentages of posts, hours or deliveries separate from the percentage of sales used in the customer review.

Then describe the price decision in ordinary language. Record who currently decides the amount, what supports that understanding & anything that remains unclear. Keep the proposal specific to the product or service at hand. The survey used a general price question; a real worksheet can distinguish arrangements for different assignments rather than assume one answer covers everything you sell.

Use the copyable fields below to prepare the discussion. They leave room for an unknown answer. A question for the client is more useful than a confident entry based only on the size of the account.

Customer & period: [customer, month reviewed, records used].

Monthly sales mix: [sales to this customer] ÷ [total sales measured on the same basis] × 100 = [share]. Missing or provisional amounts: [details].

Current price decision: [who decides the price for this product or service, how the process works & evidence or confirmation]. Open questions: [details].

Proposed work & amount: [deliverables, quantity, revisions or other scope, proposed price & currency, intended start date].

Discussion to prepare: [the change being proposed, the person to discuss it with & the question that needs an answer].

Record after the discussion: [what was agreed, what remains open & where the updated proposal is kept]. Customer-review date: [date].

For the fictional photography studio, the worksheet would connect the 60% monthly share to an owner-prepared package proposal. The discussion field could ask whether the client wants the same volume at the proposed amount or a different scope. For the second fictional provider, it could identify which part of a client-specified assignment price is open for discussion. Neither entry assumes the client will agree.

If the price process is unclear, ask a concrete question: “For the next set of deliverables, should I propose a package price, or is there an amount you want the scope to fit?” Record the answer alongside the proposed work. That gives the next draft a specific question to resolve, rather than a conclusion drawn from the size of the account.

If the review prompts you to seek additional customers, give that marketing assignment its own brief. Identify the service, the buyers you want to reach & the information you need to evaluate an inquiry. Keep that customer-acquisition goal beside the price proposal, with each document naming the specific question it is meant to address.

Retain the completed review with the proposal so that a later conversation can use the actual terms, rather than reconstructing them from a customer percentage. A sales share describes the mix for its stated period. A proposal describes an offer. Keeping both visible prevents one from being used as a substitute for the other.

Of 6,763 respondents eligible for these questions, 6,752 had both answers available. The 11 incomplete cases were kept out of the 2 reported groups rather than counted as “no.” Complete answers covered 99.8% of eligible people in the sample & 99.7% after applying survey weights.

We also checked how much the client-share difference could move if unanswered items were assigned differently while keeping the released weights fixed. A conservative outer range for the difference was 16.3–18.7 percentage points, rounded outward to 1 decimal. Under those assignment rules, the missing answers did not reverse the positive difference in the point estimates.^[3]

That range addresses the unanswered items in the released sample. It is separate from the 13.4–21.0-point sampling confidence interval & should not be read as another 95% interval. The package gives the unrounded bounds & assignment rules; the article’s main estimates continue to describe the complete-answer groups.

We used the official Full data file for Chile’s 2025 VIII Microentrepreneurship Survey, produced by the Instituto Nacional de Estadísticas (INE) with the Ministry of Economy. Its household-based design includes people working for themselves & employers operating with up to 10 workers, including the owner, in formal & informal activity. The released file contained 7,170 survey records.^[1][4]

We followed the questionnaire’s route for people continuing independent activity, including its provision for highly unstable independent activity. This left 6,763 eligible respondents. Requiring an explicit yes or no to the monthly concentration question & one of the 5 price-decision answers left 6,752. “Don’t know,” nonresponse & questions outside a respondent’s route were not converted to a substantive answer. We did not infer customer identities or add a creator subgroup.

Within each group, we used the released final survey weights to estimate the 5 actor shares. The full original strata & primary sampling units remained in the variance calculation, including units contributing zero to a particular group. We used a first-stage, with-replacement Taylor approximation based on those released design fields. This accounts for their clustering & stratification; it does not reconstruct every phase or calibration step of the original survey design.^[3][4]

Before calculating the outcomes, we specified the client-share difference as the single primary comparison. Its 95% t interval used the full covariance & the smaller group’s 631 degrees of freedom. Each actor share received a separate pointwise 95% logit-t interval using its group’s degrees of freedom. These intervals do not provide simultaneous 95% coverage for the whole table. All 10 shares met the prespecified sample-support & precision checks documented in the package.

Earlier Chilean producer work had already examined customer concentration & business decisions. This analysis connects the 2025 answers in the conditional distributions shown here.^[5] The derived tables use INE data under CC BY-SA 4.0; Influence IQ performed the calculations & interpretations. No INE endorsement is implied. The download provides aggregate results & original reproduction code, with no individual survey records or identifiers.^[3][6]