Turn market signals into your next move.

See current clients and new potential clients across the market, then move from county context to the accounts that need attention. Sales figures are illustrative sample data.

Panel mode

Click any county for the account-executive view: the existing book, budget attainment, wallet share and each client’s Keep/Grow posture.

Color counties by

–––

Filter the book 17 counties

Get / Keep / Grow — how & why

Strategy lives on the accounts, not the map. A county does not have a posture — an AE does not work a polygon, they work a book, and two accounts on the same street can need opposite moves. The affluent county with a 78%-wallet-share anchor client and a 31%-share challenger needs Keep on one and Grow on the other; scoring the county averages that away and tells the AE nothing they can act on Monday morning. So: Keep and Grow classify clients. Get classifies prospects. Geography carries neither.

GROW and KEEP — one label per client, in the AE Interface.
Three factors vote. Each returns a signed strength between −1 and +1: positive argues Grow (there is room and reason to expand this account), negative argues Keep (defend it). The votes are weighted and summed, in this priority order:
• Share of wallet — 55%. SR spend as a share of the client’s total ad budget. Low share means headroom. The ~40% rule of thumb sits at the crossover of a smooth ramp, not in an if-statement — a client at 41% is barely Keep, and one strong secondary signal flips it. Nothing snaps at a threshold.
• Projected churn — 30%. High projected churn pulls toward Keep: protect the account before pushing more product into it. It is a forward-looking projection, not history.
• YoY growth — 15%. Momentum reinforces the other two. Lightest by design — it defers rather than overrides.
Anchors are absolute, not percentiles of the current book. When real MR&I data lands the Grow/Keep split should move, and that movement is information rather than an artefact of re-normalising. Read the score, not just the label — hover any pill. A score near zero means the account is genuinely ambiguous and the AE’s judgement beats the tool’s.

GET — every tracked business that is not already a client, in the Prospecting Tool.
The question is not who is a Get (everyone not on the book is) but who to work first. Priority = how good the ground is × how good the account is, weighted 50/50:
• County tailwind — 50%. The premise: growing economies with a deep business base put more businesses in the market for ad dollars. Six inputs, all real, min–max normalised across the 17 counties — population growth 20% (Census PEP 2020–24), business base growth 20% (CBP establishments 2019–23), income growth 15% (ACS), younger skew 15% (ACS median age), business density 20% (ad-buying establishments per 1,000 residents), vertical variety 10%.
• Account fit — 50%. The existing propensity score: account size, overlap between the channels the business already buys and the channels SR sells, and how switchable the current budget holder is.
A prospect is flagged priority if it lands in the top quartile DMA-wide, not top-quartile within its own county. That is deliberate: the whole point of the tailwind term is that a mediocre account on strong ground can outrank a good account on weak ground. Ranking inside each county would throw that away and hand every county the same number of flags whether or not it is worth working.

Why vertical variety is only 10%. Variety was the original hypothesis for “diverse business ecosystem”, and it is measured here honestly — real establishment counts by NAICS sector from County Business Patterns, not from the sample prospect file, which would have measured nothing but its own random seed. It turns out to be nearly inert in this DMA. Sector evenness runs 0.87–0.92 in sixteen of the seventeen counties, with only Holmes (0.79, construction-heavy) separating — a coefficient of variation of 3%. Industry mix is close to scale-invariant across a metro like this: Carroll County has proportionally about the same retail / health / professional spread as Cuyahoga, just less of everything. Normalising a 3%-CV input to 0–100 and weighting it heavily would manufacture signal out of rounding error. The concept is right, so it stays in at the weight it actually earns; the work is carried by business density (CV 14%) and business base growth (range −2.4% to +8.1%), which measure the same idea — how much local commerce is there, and is it growing — with real spread.
The ad-buying sector restriction matters too: a county diverse in agriculture and mining is diverse, but it is not a mix of businesses buying local ads. Variety and density both count only the 11 sectors this tool’s verticals map onto (construction, retail, information, finance, real estate, professional services, education, health care, arts, food service, and other services incl. auto repair).

Caveats. Client and prospect figures are fictional sample data, so every Keep/Grow label and the fit half of every Get score is illustrative — the logic is production-ready, the numbers are not. The county tailwind underneath Get is real. The 55/30/15 posture weights and the 50/50 Get split are analytical judgment pending validation against internal performance data. Income growth compares overlapping ACS vintages, so it is directional. With 17 counties, min–max normalisation stays sensitive to outliers.

About & methodology

Why the map is built on counties. The map answers a question about market, not delivery. AEs sell CTV, streaming audio, digital display, social and search across the entire DMA regardless of cable-wire coverage, and linear reaches non-wired ground through the interconnect — so drawing the sellable world as a wire map would understate the addressable market.
The base geography is the 17 counties of the Cleveland–Akron (Canton) Nielsen DMA: Ashland, Ashtabula, Carroll, Cuyahoga, Erie, Geauga, Holmes, Huron, Lake, Lorain, Medina, Portage, Richland, Stark, Summit, Tuscarawas, Wayne. Counties are non-overlapping, they cover the market completely, they are the unit every syndicated source (Scarborough, BIA, Vivvix, IBISWorld, CBP) already cuts by, and they are a unit an AE can actually be assigned.

Strategy scoring sits on the accounts, not the county. Client-level Keep/Grow and prospect-level Get are more actionable than a single blended county score, and they don’t average opposite situations into a middling number. This is documented in full under Get / Keep / Grow — how & why above.

Geography. County boundaries are the Census 2023 cartographic boundary file (cb_2023_us_county_500k, NAD83), simplified to roughly 80m and rounded to 5 decimal places — visually identical at DMA zoom, and the whole 17-county layer costs 21 KB. Nothing is drawn by hand.

Demographics — read this before quoting a number in the room. All demographics are ACS 2020–2024 5-year, computed at county level. They will not tie to the Market Map 2026 deck. The deck’s figures are TruAudience, published per zone only; re-cutting them to counties would mean inventing a weighted average, which is exactly the fabrication this tool has refused throughout. Census is the only source that reports natively at county geography, so it is the only internally consistent choice once zones are gone. The visible difference is median age: the deck reads ~56 per zone, Census reads ~41 DMA-wide, because the deck’s figure is on a different base (adult or head-of-household, not whole population). Income, education and homeownership track much more closely. If someone has the deck open next to this tool, that is the discrepancy they will find.
• Population growth is Census PEP 2020→2024 — annual estimates, clean non-overlapping endpoints.
• Income growth compares ACS 2017–21 to ACS 2020–24 medians. The vintages overlap by two years, so it is directional only, not a clean delta.
• Business density, business base growth and vertical variety are County Business Patterns 2023 and 2019, restricted to the 11 locally-advertising sectors.
• Hispanic population index is county Hispanic share indexed to the 17-county DMA aggregate = 100, so 138 means Cuyahoga is 38% above the DMA rate — not a national index.
• Unemployment is ACS (unemployed / civilian labor force), so that every demographic on the map comes from one source at one geography.

Sample data. Every client, prospect, revenue and budget figure in this tool is fictional and flagged SAMPLE DATA wherever it appears. The 211 sample clients and 261 sample prospects were redistributed from the old zones to counties by seeded draw from each zone’s real ZIP→county distribution, so an Akron book lands in Summit rather than at random. The prospect universe contains the client book — clients appear with status “Client” and their SR revenue — so captured + whitespace = addressable and the two panel modes reconcile county by county. County budgets were split from the old zone targets in proportion to where each zone’s book landed. A real build replaces all of it with CRM and Sales Ops extracts joined to a competitive-spend source (Vivvix/Kantar) and a business file (D&B, Data Axle). What is real: all geography, all demographics, and the entire county tailwind that drives half of every Get score.

Panel modes. The AE Interface is the inward view — the existing book in a county, budget attainment, wallet share, and each client’s Keep/Grow posture. The Prospecting Tool is the outward view — the county tailwind, every tracked business whether or not it is a client, addressable spend, whitespace, and ranked Get targets. Toggling modes never changes the map: identical counties, colors, filters and selection; only the right-hand panel re-renders.

Budget & sales plan. Each county carries an annual revenue target. The AE panel shows budget vs. booked, an attainment bar and the variance — green over plan, red under — measured on the full book, so it does not move when filters change. Any county under plan gets a Sales plan: its open accounts in Get-priority order, each sized at open whitespace × the county’s own share of wallet × close probability (10–45%, from the fit score), stacked until weighted pipeline reaches 2.5× the gap — a coverage ratio, because most opportunities don’t close. Where a county’s whole prospect file can’t carry that, the panel says so rather than inventing a plan. The close-rate curve and the 2.5× coverage ratio are assumptions pending calibration against real win rates.

Color scale. Green = favorable, red = unfavorable, and each metric declares a direction so “good” is always green (low unemployment maps to green). Purely descriptive demographics — median age, households with children, homeownership, race/ethnicity, build year — use a neutral light-to-dark blue ramp instead: painting population attributes as good or bad would be misleading, and a bad look in an SLT room. Nothing on the map is a strategy score.

Basemap © OpenStreetMap contributors © CARTO.

This visualizer was created by George Lee.