Back to Blog
BlogUnited States8 min read

United States · US

Run a small territory pilot that explains why conversations stall

Use a real customer-discovery case, federal research guidance and clearly defined account records to design a small territory pilot that explains what to change next.

By the playzii team

Updated on

Report scope: fiscal years 2023–2024
Authentic original text excerpt. Program descriptions are not evidence of a territory pilot's sales results.

Courtesy: U.S. National Science Foundation, 2025 I-Corps biennial report. Captured and cropped by playzii, October 26, 2026.Image source · Public-domain federal agency explanatory text, unmarked as third-party copyright. NSF text-reuse policy https://www.nsf.gov/policies/digital explicitly permits copying this text; separate visual-media permission policy respected by excluding agency logos, photographs, illustrations, figures/charts and all third-party art. Direct live-page screenshots of text only, no reconstructed typography or endorsement.

Make the pilot answer a useful question

A territory pilot should help a team make its next decision: keep the account definition, change the problem it investigates, or stop spending time on an unsuitable group. A spreadsheet full of activity cannot answer those questions by itself. Before contacting companies, decide what evidence would distinguish a useful conversation from a polite response and what you will record when someone declines.

This guide is for US B2B firms testing a service in a limited territory. It combines federal market-research guidance, a customer-discovery program report and an actual aircraft-development case. The proposed workflow is an editorial adaptation for a small sales team. It has no prescribed sample size, promised conversion rate or claim that a few interviews represent a regional market.

State the question before choosing the accounts

Start with a question that can change what you do. For example, a maintenance provider might ask whether independent plants within its service radius struggle to arrange a particular inspection. That is more useful than asking whether manufacturers like the company. Write down the problem, the relevant operating setting and what the team could realistically deliver before collecting names.

The SBA's market-research guidance distinguishes existing research from direct research and encourages examining demand, location, alternatives and pricing. Use broad evidence to define a reasonable starting group, then ask about actual work. Published industry size cannot tell you whether an individual plant has a problem worth discussing this month.

Make eligibility independent of willingness to answer

Define which accounts belong before you know who replies. Eligibility might depend on an operating location, a relevant process and the ability to serve the site. Record why each company meets those conditions. A large employer in the right county can still be unsuitable if the relevant work happens elsewhere or the service is already handled by a parent organization.

Keep unresolved cases visible. An account whose process is unknown is different from one confirmed to be outside the scope. Quietly removing difficult cases after seeing poor responses makes the pilot look more promising without improving the account definition. Equally, adding friendly existing customers halfway through creates a different group. Record such changes and review their results separately.

Use regional statistics with their dates attached

County Business Patterns can provide context for a geography and industry. The Census program description identifies establishments with paid employees, rather than a list of individual buying organizations. At the research check, that page also warned that its information was no longer current while the agency evaluated alternatives to noise infusion. Treat that warning as part of the source, not as a footnote to ignore.

The retained 2023 CBP data user guide, updated July 16, 2025, says the program does not provide company lists or sales and receipts. It is historical documentation, not proof of the next release's methods. A territory pilot therefore needs its own verified account research, and any numerical comparison should retain the dataset year, industry classification and applicable methodological notice.

Customer discovery and testing potential commercial applications
Authentic original text excerpt. Program descriptions are not evidence of a territory pilot's sales results.

Courtesy: U.S. National Science Foundation, 2025 I-Corps biennial report. Captured and cropped by playzii, October 26, 2026.Image source · Public-domain federal agency explanatory text, unmarked as third-party copyright. NSF text-reuse policy https://www.nsf.gov/policies/digital explicitly permits copying this text; separate visual-media permission policy respected by excluding agency logos, photographs, illustrations, figures/charts and all third-party art. Direct live-page screenshots of text only, no reconstructed typography or endorsement.

A discovery program offers a useful distinction

The 2025 NSF I-Corps biennial report, published in March 2026, describes learning about commercial applications through customer discovery. Its national program trained 509 teams in fiscal years 2017–2018 and 685 in 2023–2024, with the intervening totals shown below. These are program participation counts. They do not measure customers acquired or prove that interviews caused business success.

The national program solicitation expects at least 100 interviews with potential customers and partners. That requirement belongs to a particular funded program. It is not a statistically justified minimum for your territory pilot. Borrow the discipline of testing assumptions while setting an amount of work your own team can complete and review carefully.

A real case shows why the question can change

The Great Plains hub's January 2026 Ilios Aerospace case describes interviews changing the team's assumptions about aircraft features. The team reported learning that some prospective customers did not want an amphibious configuration, leading it to consider modular floats. The account reports 105 interviews in five weeks, beyond its regional program's stated goal of 20.

This is a program participant's experience reported by its support organization. It is not a controlled evaluation, a sample of US industrial buyers or evidence of certified aircraft sales. Its practical lesson is narrower: record feedback that challenges a favored feature. A pilot that only collects endorsements cannot reveal whether the offer itself needs to change.

National team roles: entrepreneurial lead, technical lead and industry mentor
Authentic original text excerpt. Program descriptions are not evidence of a territory pilot's sales results.

Courtesy: U.S. National Science Foundation, 2025 I-Corps biennial report. Captured and cropped by playzii, October 26, 2026.Image source · Public-domain federal agency explanatory text, unmarked as third-party copyright. NSF text-reuse policy https://www.nsf.gov/policies/digital explicitly permits copying this text; separate visual-media permission policy respected by excluding agency logos, photographs, illustrations, figures/charts and all third-party art. Direct live-page screenshots of text only, no reconstructed typography or endorsement.

Give each conversation a clear purpose

Invite the person to discuss a relevant work problem and explain how you intend to use the conversation. Ask about a recent instance, who handled it, what happened next and which alternatives were available. Someone describing an inconvenience is giving you evidence about a process. They have not necessarily asked for a proposal or committed their organization to buying anything.

Do not disguise a sales pitch as independent research. If the conversation moves toward your service, acknowledge that change and ask whether discussing a possible next step would be useful. If you use commercial email to recruit participants, review the FTC's business guidance. The agency explicitly includes B2B commercial email. A small pilot is still real outreach and needs truthful messages and a functioning way to stop further marketing contact.

Track accounts and contacts separately

An account may have several relevant people, and one person may discuss more than one operating site. Count eligible accounts, people approached, replies, completed conversations and agreed next steps separately. Otherwise, three helpful interviews at one business can appear to be evidence from three different buyers. Record the role and site relevant to the statement without collecting unrelated personal information.

Also record why a conversation could not happen. A wrong role, an invalid address, a refusal, an unavailable person and an unanswered invitation call for different decisions. None of them proves the underlying business has no need. This distinction gives the team something concrete to improve without inventing motives for people who did not respond.

Keep the offer stable enough to interpret

Use the same central problem question across the initial group, while allowing people to describe their own work. Record meaningful changes to the introduction, eligibility criteria or proposed service. If a pilot simultaneously changes its territory, message, price and contact channel, it becomes hard to know which part deserves further investigation.

That does not mean repeating a bad approach until the calendar ends. Correct an obvious error, explain the change in the record and review earlier conversations separately where necessary. An operational learning exercise can be useful without being a randomized experiment. Be honest about what was changed and avoid presenting the result as a causal comparison between messages.

The report describes institutions and mentors supporting a hub
Authentic original text excerpt. Program descriptions are not evidence of a territory pilot's sales results.

Courtesy: U.S. National Science Foundation, 2025 I-Corps biennial report. Captured and cropped by playzii, October 26, 2026.Image source · Public-domain federal agency explanatory text, unmarked as third-party copyright. NSF text-reuse policy https://www.nsf.gov/policies/digital explicitly permits copying this text; separate visual-media permission policy respected by excluding agency logos, photographs, illustrations, figures/charts and all third-party art. Direct live-page screenshots of text only, no reconstructed typography or endorsement.

Review objections as evidence about different decisions

When someone declines, preserve the reason they actually gave. A company outside the service radius, a team with an existing contract, a manager unable to release staff time and a buyer who does not trust your evidence have different concerns. Do not compress them into a single price objection simply because price is the easiest field to fill.

For each recurring issue, decide what would answer it. Coverage may require a different territory. A claim about performance may require a relevant case with a clear measurement period. An existing contract may justify a later conversation only if the person agrees. Some objections reveal a poor fit that no additional message should try to overcome.

Choose the next step before calculating a rate

A useful next step is specific enough for both parties to recognize: a discussion with the site engineer, permission to review a relevant document or a request for a scoped assessment. Keep proposals requested, proposals delivered and orders separate. A friendly reply should not move an account into the same state as an approved purchase.

When reporting a proportion, state the underlying counts and the observation window. Small, self-selected groups can change sharply with one additional reply. Comparing two territories that differ in eligibility, contact effort or existing relationships does not isolate a territory effect. Review those differences before deciding that one region is more receptive.

End with a decision the team can act on

At the review, read examples alongside the activity counts. Ask whether the original problem appeared in relevant accounts, whether the right roles participated, what prevented progress and what the company would need to deliver responsibly. Include cases that contradict the preferred explanation. A pilot should make an unsuitable offer easier to abandon, not only make promising results easier to display.

Choose a concrete next action and its owner. The team may narrow the operating situation, change the evidence it brings, conduct a further set of interviews or stop the pilot. Keep account and contact notes linked to the original facts and agreed next steps so another colleague can understand the decision. There is little value in a successful meeting count that nobody can explain a month later.

Regional and national programs have different participation requirements
Authentic original text excerpt. Program descriptions are not evidence of a territory pilot's sales results.

Courtesy: U.S. National Science Foundation, 2025 I-Corps biennial report. Captured and cropped by playzii, October 26, 2026.Image source · Public-domain federal agency explanatory text, unmarked as third-party copyright. NSF text-reuse policy https://www.nsf.gov/policies/digital explicitly permits copying this text; separate visual-media permission policy respected by excluding agency logos, photographs, illustrations, figures/charts and all third-party art. Direct live-page screenshots of text only, no reconstructed typography or endorsement.

Sources, screenshots and limits

Research was checked on October 26, 2026. The NSF report covers fiscal years 2023–2024 and reports national and regional programs separately. This article uses only national team counts in its chart; it does not combine startups, individuals and teams. The Ilios case remains an attributed account, and the example maintenance pilot is hypothetical. No outcome from a playzii customer or live outreach campaign is asserted.

The five images are genuine, distinct captures of NSF's own explanatory text, with linked credits. They show the report period, customer-discovery description, team roles, hub structure and regional-program differences. NSF's text-reuse policy permits copying unmarked public-domain text and treats visual media separately. We excluded its photos, logos, illustrations, charts and third-party material. Ordinary crops and enlargement improve readability without reconstructing the page. The proposed workflow and new chart are editorial work; NSF does not endorse them.

Research data

National I-Corps teams trained in successive two-year periods

Counts in Table 1 of the 2025 NSF report. These are national-program team participation counts, not individual interview counts, startups formed, customers acquired or a causal effectiveness estimate.

National teams trained · 2024-09-30
National I-Corps teams trained in successive two-year periodsCounts in Table 1 of the 2025 NSF report. These are national-program team participation counts, not individual interview counts, startups formed, customers acquired or a causal effectiveness estimate.. FY2017–2018: 509 National teams trained. FY2019–2020: 593 National teams trained. FY2021–2022: 638 National teams trained. FY2023–2024: 685 National teams trained.FY2017–2018509FY2019–2020593FY2021–2022638FY2023–2024685

Transcribe the national teams trained row for four fiscal-year pairs. Do not add the cumulative since 2012 column or compare regional-program counts as though populations were identical. Report published March 16, 2026, updated April 9, 2026; period ends FY 2024.

Sources: U.S. National Science Foundation

View data and sources
National I-Corps teams trained in successive two-year periods · National teams trained
IndicatorValueSource
FY2017–2018509 National teams trainedU.S. National Science Foundation: 2025 NSF Innovation Corps biennial report
FY2019–2020593 National teams trainedU.S. National Science Foundation: 2025 NSF Innovation Corps biennial report
FY2021–2022638 National teams trainedU.S. National Science Foundation: 2025 NSF Innovation Corps biennial report
FY2023–2024685 National teams trainedU.S. National Science Foundation: 2025 NSF Innovation Corps biennial report

Sources and references

  1. 2025 NSF Innovation Corps biennial reportU.S. National Science Foundation · US

    Accessed

  2. NSF 25-549 National Innovation Corps Teams solicitationU.S. National Science Foundation · US

    Accessed

  3. Accessed

  4. Accessed

  5. Accessed

  6. Plan your business: market research and competitive analysisU.S. Small Business Administration · US

    Accessed

  7. CAN-SPAM Act: compliance guide for businessFederal Trade Commission · US

    Accessed

Keep account research and next steps together Back to Blog