> ## Documentation Index > Fetch the complete documentation index at: https://course.pokesignal.io/llms.txt > Use this file to discover all available pages before exploring further. # 1.4 ยท Run a Real Sales Test > Design a small sales experiment that measures the recurring work, looks for a repeated constraint, and ends in a continue, change, or stop decision. Use the resources you recorded in [Chapter 1.3](/chapters/1-3-the-three-ingredients) to run a small batch of real sales. Aim for roughly twenty completed orders before adding optional tools or capacity. Twenty is not a magic threshold; it is a practical amount of repetition. The point is to encounter the same work more than once and see whether a pattern appears. This experiment answers three questions: Can your current resources complete the selling loop? Where does the work repeatedly slow down or fail? Do you want to keep doing the recurring parts after the novelty fades? ## Set the boundary before you list Keep the test narrow enough that you can explain what happened. **Inventory scope:** \_\_\_ Choose cards you have already cleared for sale and can identify and condition with reasonable confidence. **Provisional venue for this work-fit test:** \_\_\_ Use one venue that can carry the batch without unnecessary setup. This chapter tests whether you can perform the recurring work; it does not choose the permanent or best sales venue. Keep the venue, cohort, order, time, and friction records. [Chapter 4.2](/chapters/4-2-choosing-where-to-sell) will use them with the Part 2 economics to confirm or change the primary venue. **Target:** Roughly twenty completed orders. If the review date arrives first, review the sales you completed and record what prevented the rest. **Start date:** \_\_\_ **Review date:** \_\_\_ Leave enough time for ordinary work and customer contact to appear. **Maximum owner hours:** \_\_\_ hours **Maximum additional cash:** \$\_\_\_ **Maximum learning loss from my first-year memo:** \$\_\_\_ Copy this boundary from [Chapter 1.2](/chapters/1-2-what-success-actually-looks-like). Part 2 will define the economic measure before the final review; unsold inventory is not automatically a realized loss. Bring these boundaries from the decisions you made in the previous chapters. **Continue if this observable result occurs:** \_\_\_ **Change if this observable result occurs:** \_\_\_ **Stop or pause if this observable result occurs:** \_\_\_ Write the decision rule before the results are known. Use only the supplies and systems needed to finish this batch. A purchase that makes the experiment possible is different from a purchase intended to make a future, larger operation faster. ## Count the complete loop A listing is not a completed sale. For this experiment, count an order after it has been paid, prepared, and fulfilled. Continue recording related customer contact until the review date. The loop begins before the listing. You may need to identify the exact card, assess condition, check a realistic value, create the listing, locate the sold item, pack it, and resolve a question or mistake. Part 4 teaches those workflows in detail. Here, you are observing where your first version holds together. Do not optimize every step while the test is running. Correct errors that affect a customer or the accuracy of the experiment, but resist changing the venue, inventory type, schedule, and tools all at once. If several variables move together, the result will be hard to interpret. ## Record friction, not impressions Keep one small operating log. A row for every order is unnecessary; update the same four observations as the batch develops. **Log update checkpoint:** after each \_\_\_ | Observation | What to record | Pattern so far | Response to test | | --------------------- | ----------------------------------- | -------------- | ---------------- | | Owner time | Total and stage using the most time | \_\_\_ | \_\_\_ | | Errors or corrections | Type and count | \_\_\_ | \_\_\_ | | Customer contacts | Reason and count | \_\_\_ | \_\_\_ | | Work waiting | Stage and duration | \_\_\_ | \_\_\_ | Be concrete. *Listing was annoying* is an impression. *The second half of the batch waited four days because I had not finished condition checks* identifies a stage, a delay, and a possible constraint. An isolated mishap may need a simple correction. Repeated friction deserves a closer look. The response may be a narrower batch, a clearer procedure, a skill you need to practice, a change in schedule, or eventually a tool. Do not assume equipment is the answer before the experiment shows what is actually limiting the work. ## Debrief the pattern At the review date, bring forward the completed orders, unfinished inventory, and four observations from your operating log. Add the recurring task you tolerated least and the one you handled well. If one problem repeated, name that constraint. If no clear pattern appeared, record that result instead of promoting an isolated mishap into a diagnosis. A small batch should produce the narrowest conclusion the records support, not the longest list of things that could someday be improved. Compare the result with the boundaries you set. Did the batch fit the available time and cash? Could you complete the work accurately? Did the recurring work feel acceptable once it stopped being new? Your records should make those answers more specific than a general feeling that the test went well or badly. ## Make the decision you wrote **Continue** when the loop fits the stated limits and no recurring problem makes the current version unworkable. **Change one variable** when the work still appears viable but the same constraint keeps returning. Change the inventory scope, schedule, procedure, or another relevant input, then run the next batch without changing everything else. **Stop or pause** when the test breaks a boundary you are unwilling to move, or when the recurring work itself is not something you want to continue. The experiment does not prove that the business will scale. It gives you something more useful at this stage: direct evidence about your own selling loop and one informed next decision. Part 2 now gives you the economic model needed to judge the inventory and orders inside that loop.