Competitive Brief · Prospect Research
DonorAtlas
An AI-native donor prospect-research product built on “bottom-up net worth.” It is, in effect, the productized version of what our deeds + bios + donor + prospects verticals assemble — minus the SEC layer that is our edge.
Prepared for Ronda, Elise, Robin, Raul · 2026-08-19 · from Ronda's Apra FL webinar screenshots + public sources
At a glance
Founded2023
HQNew York, NY
Team~7 people
Raised$3M
Coverage claim200M+ US people
FounderWill Schrepferman
Investors: 1984 Ventures, Audacious Ventures, Higher Ground Labs, Soma Capital. Sells annual licenses (quote-based, no public pricing). Integrates with Blackbaud Raiser's Edge NXT / ResearchPoint and Salesforce. Buyers: nonprofits, higher-ed, healthcare, political campaigns, fundraising consultants.
The mirror test: we already hold DonorAtlas in our graph
The sharpest proof of the difference is DonorAtlas itself — it's a node in our own graph (company_id 741277b0…), and we hold the private-company detail its bottom-up model can only estimate:
Round$2.75M Seed
DateSep 4, 2024
Tiervouched
FounderWill Schrepferman
Investors on the round (named): 1984 Ventures, Audacious Ventures, Higher Ground Labs, Soma Capital.
Recall their own founder-profile screenshot had to guess his equity from a press-estimated valuation. We hold the dated round and the full investor set — filed, not inferred. The competitor is a worked example of the exact private-company / cap-table intel a web-scrape + property model structurally cannot produce.
The pitch: bottom-up vs. top-down
Their entire wedge is a shot at the incumbents (WealthEngine, DonorSearch, iWave):
- “Legacy” top-down modeling — heavily property-based, modeled from aggregate population statistics, and (their words) “fails at HNW + UHNW.” They also reject “black-box scores.”
- DonorAtlas bottom-up — assemble a net-worth range from observable components (income, real estate, business/foundation assets) with an AI narrative and clickable citations on every figure, refreshed from the open web.
What the product actually does
From the live profile screenshots (a person named “Cristi Pittman,” and their own founder's profile):
- Estimated Net Worth as a range on a log slider ($3.4M–4.9M) + Estimated Liquidity (liquid↔illiquid).
- Predicted Annual Giving ($5.9K) and cause-affinity tags (“Top Issues”: Education, Scholarships, Family Services…).
- AI bio with numbered citations — career, education, spouse, family, professional fee ranges — every claim linked to its source.
- Income table (employer, role, dated salary range) + Education.
- Deep property / deed panel: est. value, current deed owners, equity vs. mortgage balance, and full deed history incl. intra-family transfers ($0 sale price).
- Relationship graph (spouse/family, “connections to your network”) + a “Donors like ___” lookalike recommender.
- Private-equity / startup-equity inference — it estimated the founder's own stake from press about the company's valuation.
- Contact info + address history, natural-language search, batch wealth screening, Salesforce / Raiser's Edge sync, PDF + shareable reports.
Where the data comes from
Cited sources visible in-product: LinkedIn, Facebook, Avvo / Lawyers.com / review sites, board-member listings, property & deed records, news, nonprofit 990 filings, company databases. Note what is absent: SEC. No Form 4 insider sales, 13D/G, Form D private rounds, or fund AUM anywhere in the deck.
Head-to-head vs. Kaleidoscope / MARS
| Capability | DonorAtlas | Us today |
|---|
| Net-worth range + liquidity model | Packaged, calibrated | Components exist (property + income + holdings); not packaged as a calibrated estimator |
| Property / deed depth (equity, mortgage, deed history) | Strong, national | deeds vertical — early, coverage-limited |
| AI cited bio / employment / education | Strong | bios vertical — comparable approach (Qwen + citations) |
| Relationship / network graph | Shallow — immediate family + a single-hop “connections to your network” match + lookalikes | Deep, multi-hop, traversable — 18M board seats, 11M insider links, 1.5M jobs, 305K alumni, co-investors, M&A, family, and foundation boards |
| Predicted giving + cause affinity | Yes (modeled) | Not built (need FEC + 990 giving history + model) |
| Verified consumer contact (email/phone) | Yes | SEC-derived addresses only; no consumer email/phone |
| Social (Facebook, etc.) | Yes | No — and not planned |
| SEC insider / private-market wealth events | None | Deep + verified + dated (Form 4, 13D/G, Form D, 144, fund AUM) |
| End-user product surface (UI, CRM sync, PDF) | Full product | MCP + substrate by design; no packaged UI |
| Coverage breadth | ~200M US people | ~1M+ entity-resolved, richest on SEC/deal-side people |
The one-liner: they win on property + social + packaged UX + giving model; we win on two things they structurally lack — SEC / insider / private-market signal (verified, dated) and a deep, traversable relationship graph (family → foundation → board → co-investors, multi-hop). Two tells: their founder-profile screenshot guessed his startup's valuation while we hold the actual Form D ($3M raised); and a Tampa-lawyer search returns one record on their side vs. a full family + foundation + board network on ours.
Gaps we'd need to close if we chose to compete
Grouped by honesty about whether we'd actually do it.
Won't / can't easily
- Facebook & broad social data — not pursuing (ToS/scraping exposure; they likely have a source we can't cleanly replicate). Accept this as a permanent gap.
- Verified consumer email/phone at scale — that's a buy, not a build (the ZoomInfo-space problem; see task #320). SEC gives us mailing addresses, not personal cell numbers.
- 200M-person general-population coverage — reaching every mid-tier donor means licensing consumer data. Our depth is on the SEC/wealth-event population, not the whole donor pyramid.
Could close relatively cheaply — we're already building it
- Property / deed depth — the deeds vertical is exactly this; the gap is coverage + a property-data source, not method.
- Predicted giving + cause affinity — FEC political giving + 990 nonprofit giving + a simple model. We already have FEC in the deeds roadmap.
- Net-worth range estimator — assemble property equity + income + insider/private holdings into a calibrated range + liquidity split.
- Bio / employment / education / lookalikes — bios + person_edges already do this shape.
Our edge they can't easily close
- A deep, traversable relationship graph — this is the one that shows up loudest in a live test. Searching a Tampa lawyer on DonorAtlas returns essentially one record; on our graph the same person opens into his whole family network, the foundations he and his family run, who else sits on those boards, his companies, co-investors, and alumni ties — multi-hop, not a single “people you may know.” For major-gift work — “how is this prospect connected to our board, and who else should we be talking to?” — that is the product. Powered by 18M board seats, 11M insider links, 305K alumni, plus foundation-board and co-investor edges.
- SEC insider & private-market wealth events — Form 4 stock sales, 13D/G activist stakes, Form D private rounds, Form 144 planned sales, 13F/AUM. Verified, dated, and invisible to a web-scrape + property model. This is the “she isn't understanding SEC” gap the prospect flagged — a feature, not a liability, once made legible. It also feeds the graph: the verified board/officer/ownership edges the graph traverses come largely from these filings, which is exactly why a web-scrape can't reproduce it.
- Entity resolution — CIK/CRD/ticker-anchored canonical people & companies, so a wealth signal and a network edge attach to the right person.
Strategic options
1. Compete head-on — build the product surface (UI + CRM sync + PDF + net-worth/giving models). Biggest lift is the packaged app, not the data. 2. Differentiate on SEC + license the signal — sell our insider/private-market wealth layer into DonorAtlas-class tools that can't build it (the task #320 play). 3. Partner — DonorAtlas already bolts onto Blackbaud ResearchPoint; an SEC-signal feed could bolt onto them the same way.
Lean: we don't win a UX race against a VC-funded product team; we win by owning the two things they structurally lack — the SEC/insider signal and a deep, traversable relationship graph (and the two compound: SEC filings are what make the graph's edges verified). Lead every conversation in this market with those two, demo the network expansion on a real name (the Tampa-lawyer test), and treat property/bio/giving as table-stakes we fill from the verticals already in flight.