What Is Business Attraction in Economic Development, and What Should You Have Ready at Each Stage?
Michael Etheredge is VP of Product Development at QuestionFuel, where he leads how the firm measures and corrects the way AI answer engines describe companies. He was featured by the Southern Economic Development Council on how AI is changing the way companies research communities and regions.
Business attraction is the part of economic development that recruits new industry and private investment. A company or its consultant starts with hundreds of communities and eliminates most of them before anyone visits. This guide explains the three phases of a site search, what each phase asks an economic development organization to show, and how to check what ChatGPT says about your community before a project does.
Key takeaways
- Business attraction recruits new industry and private investment, and it works differently from business retention and expansion, which supports companies already in the community.
- A site search is a process of elimination. One university source describes three phases that move from hundreds of communities to 10 to 15, then to 3 to 5 finalists, then to a negotiated decision.
- Area Development's 2026 survey reports that site readiness and due diligence status rose from 78% to 98.5% as a factor, so documented readiness now screens communities in or out.
- One consultant argues that the first screen now happens before the RFI. We found no published measurement of how often that happens, so the practical step is to test what ChatGPT says about your community.
Quick answer: Business attraction is the part of economic development that recruits new industry and private investment. A site search narrows hundreds of communities to a few finalists in three phases. An organization needs sites, utilities, workforce facts, permitting speed, and incentives ready for each phase, stated once with dates, and checked in ChatGPT.
What is business attraction in economic development?
It is the set of activities an economic development organization uses to bring new employers and private investment to a community.
A Kansas Department of Commerce training deck posted in December 2021 quotes the International Economic Development Council definition of economic development: a program, group of policies, or activity that seeks to improve a community's economic wellbeing and quality of life by creating or retaining jobs that provide growth and a stable tax base. The same deck describes business attraction as a "Focus on attracting new industry and private sector investment."
Attraction is one strategy among several. The deck lists business retention and expansion, site and building development, entrepreneurship, workforce development, and talent attraction as separate areas of activity, and it states that there is no universal checklist. Business retention and expansion works with companies already in the community. Attraction works with companies that are deciding where to go.
The deck gives attraction examples that are still concrete: mock site visits with local volunteers, current and easy to access information on sites, transportation, and utilities online, direct marketing to site selection consultants, and hosting a prospect or a site selector.
How does a site search move from hundreds of communities to one?
It moves through three phases of elimination, and each phase asks for different facts.
A University of North Carolina School of Government article, published November 7, 2025, describes a typical project in three phases. In the first, the company or its consultant reviews responses to a request for information, or RFI, and screens out communities that miss project criteria. The article says that phase may start with hundreds of communities and end with 10 to 15 or so that remain viable. In the second, remaining locations get a detailed review of community data, site and building criteria, cost modeling, and site visits, which narrows the field to 3 to 5 finalists. In the third, the company closes with one finalist after more cost and return on investment analysis, and incentive negotiations take place.
The Kansas deck draws the same funnel with slightly different counts: hundreds of possible sites, 12 to 15 options, then 2 to 3 finalist visits. Treat the counts as typical, not fixed. Each project differs.
| Phase | What happens | What the organization needs ready | Source |
|---|---|---|---|
| 1. Screening | RFI responses reviewed against project criteria; hundreds of communities reduced to about 10 to 15 | Accurate, consistent facts on sites, power, water, workforce, and permitting | UNC School of Government, November 2025 |
| 2. Detailed review | Community data, site criteria, cost modeling, site visits; 3 to 5 finalists | Site documentation, workforce data, a visit plan, responsive staff | UNC School of Government, November 2025 |
| 3. Close | Return on investment analysis and incentive negotiation | A clear financial case and incentives matched to the project's constraint | UNC School of Government, November 2025 |
Where do answer engines fit into those phases?
They may sit before Phase 1, and that is the part no one controls.
In a Q3 2026 Area Development article published July 15, 2026, consultant Devin Hillsdon-Smith of Hyphen Strategies argues that "The RFI Is No Longer the Starting Line." He writes that "AI has collapsed that timeline from weeks to hours," and that communities can be eliminated "in an afternoon" before an RFI is drafted. He lists what an assistant may read about a community: city council minutes, utility rate histories, permit timelines, environmental records, and social media sentiment.
Read that as one practitioner's account, not a measurement. The article gives no data on how many projects work this way, and we found no published study that does. The inference is still worth acting on, because checking costs little. If a site selector asked ChatGPT about your community tomorrow, you would want to know what it said.
AI Narrative Drift: the gap between what is currently true about a company and what answer engines say about it, because those answers update slowly and imperfectly when facts change.
AI Narrative Drift is the name for that gap. For a community, it could show up as an old power capacity figure, a closed program still described as open, or a missing workforce fact. The first step of the DRIFT Framework is Detect: ask, record, and compare against your own documented record.
What do site selectors say they weigh in 2026?
They say certainty, which in practice means documented readiness, speed, power, and talent.
Area Development's 2026 survey, reported in Q1 2026, frames the shift as a move from cost to risk. In its words, the question companies now ask is not "where is cheapest?" but "where can we be sure this will actually work?" The survey reports these figures:
| Factor | Reported result | What it suggests |
|---|---|---|
| Site readiness and due diligence status | 98.5%, up from 78% the year before | Documented readiness now screens communities in or out |
| Certainty of permitting | 94% of respondents | Permitting speed is a top tier factor |
| Responsive state and local government | 97% | Staff responsiveness is part of the evaluation |
| Availability of skilled labor | 100% rated important or very important | Workforce is still the first filter |
| Depth of technical talent pipelines | 95.5%, up from 79.7% | Community colleges and trade schools are now evidence |
| Electric power availability at scale | 98.5% | Power is a qualifying question |
These are survey results about what advisors say they weigh. They do not show how often ChatGPT repeats or ignores these facts. For the regional version of that question, see Does AI Describe Your Region by the Criteria Site Selectors Now Use?.
What should an economic development organization have ready at each stage?
It should be able to answer five practical questions without hesitation, and then state the answers where they can be found.
Danielle Campolo of Strategic Development Group wrote in Area Development on September 23, 2026 that her firm evaluated more than 500 sites across 28 states in 2025, and that five factors drive nearly every client's location decision, in the order in which they tend to eliminate communities. We rephrase them as a readiness list:
- •Do you have a site or building? Campolo describes a search for an existing building that went out to 17 states and returned seven candidates. Her advice is to keep a pipeline behind what is available today.
- •Can you meet logistics and utility needs? State available power capacity, water rates and reserves, and what transportation supports. If a constraint exists, say what plan, timeline, and funding address it. In her words, "Companies can work around limitations. They can't work around uncertainty."
- •Is the workforce there, and is more coming? Describe the pipeline from community colleges and technical programs, not only current headcount.
- •Do you understand your financial case? Present land costs, utility rates, taxes, and incentives. "If an approval process takes six months, that's a liability, not an incentive."
- •Do people feel wanted? Responsiveness and follow through at the site visit can break a tie.
Map them to the phases. Questions 1 to 3 carry Phase 1, when facts are screened. Questions 3 to 5 carry Phase 2 and Phase 3.
How can you check what ChatGPT says at the first stage?
Write your facts down first, then compare what ChatGPT says to them.
- •Write a one page fact record. Include sites and buildings, power and water capacity with dates, workforce programs, permitting timelines, and incentive programs, each with a checked on date.
- •Write five prompts a site selector might type. For example: "What are the main reasons to eliminate [community] for a 500 employee manufacturing facility?" Hillsdon-Smith uses a prompt of this kind as his example.
- •Run each prompt three times in new chats. One run is one sample, and answers can vary. The reasons are covered in Why Does ChatGPT Give Different Site Selection Shortlists to the Same Question?.
- •Mark every fact as a match, a gap, or an error. A match agrees with your record. A gap leaves out a fact that matters. An error states something your record does not support.
- •Save the date and the engine. Repeat after any change to the facts.
The full method is in How to Audit What AI Says About Your Region. For a single site, use How to Describe a Certified Industrial Site So Answer Engines Get It Right.
What is not known yet?
Several things, and we say so plainly.
- •We found no published study of how often site selectors use ChatGPT to screen communities.
- •We found no published measurement of how accurately ChatGPT describes individual communities on the factors in the table above.
- •Hillsdon-Smith's account of pre screening is a practitioner's view in a trade publication. The UNC School of Government article does not address answer engines at all.
- •The counts at each phase vary by project and by source.
When is this not the right approach?
It is not the right first step when the community has nothing yet to document.
The Kansas deck asks attraction teams directly whether they have "sites or buildings or just green space," and whether permitting offices and councils can "move at the speed of business." If the honest answers are no, readiness work comes before any description work. A fact record cannot make up for a missing site or a utility gap. It is also not a substitute for relationships, which Campolo and Hillsdon-Smith both say still decide close finals.
Frequently asked questions
What is the difference between business attraction and business retention and expansion?
Attraction recruits new companies from outside the community, while retention and expansion supports companies already there. The Kansas Department of Commerce deck lists them as separate strategies and uses retention visits to learn what existing companies need.
How many communities does a site search start with?
It varies by project. A UNC School of Government article says Phase 1 may start with hundreds of communities and end with 10 to 15 that remain viable. Treat that as a typical range, not a rule.
Do site selectors really use ChatGPT to screen communities?
We do not know. One consultant argues in Area Development that assistants now pre screen communities before an RFI is drafted, but the article gives no data. We found no published measurement.
What should we fix first?
Fix errors before gaps, and start with facts that screen communities out: site availability, power and water capacity, permitting timelines, and workforce pipeline. Then correct the source of the wrong fact and test again.
Sources and references
- •Kansas Department of Commerce, Office of Rural Prosperity, Economic Development 101 training deck, December 2021, checked October 23, 2026
- •University of North Carolina School of Government, Community and Economic Development, Perspectives on the Business Location and Site Selection Process, November 7, 2025, checked October 23, 2026
- •Area Development Research Desk, 40th Annual Corporate and 22nd Annual Consultant Site Selection Survey Results, Q1 2026, checked October 23, 2026
- •Danielle Campolo, Strategic Development Group, Area Development, Ready to Show: Five Things Every EDO Should Be Able to Answer Right Now, Q3 2026, published September 23, 2026, checked October 23, 2026
- •Devin Hillsdon-Smith, Hyphen Strategies, Area Development, The RFI Is No Longer the Starting Line, Q3 2026, published July 15, 2026, checked October 23, 2026
Related reading: How Do Answer Engines Describe Regions? A Guide for Economic Development Organizations · How to Audit What AI Says About Your Region · Why Does ChatGPT Give Different Site Selection Shortlists to the Same Question? · How to Describe a Certified Industrial Site So Answer Engines Get It Right