The company
Deloitte
Deloitte is one of the largest professional services firms in the world. Inside a multi-client practice, every engagement can mean a different state, county, or municipal buyer, not just a federal agency set. Capture has to cover that long tail without becoming a body shop.
Before SamSearch
SLED is not one portal. It is thousands.
State, county, and municipal work does not live on SAM.gov. It lives across 50,000+ separate SLED sources: state boards, random counties, tiny municipalities, school districts, and authorities with their own posting habits. For a consulting team chasing that long tail, portal-by-portal watching is not diligence. It is a headcount trap.
How they use SamSearch
The capture workflow
In only 6 weeks, a Deloitte team ran SamSearch as SLED capture automation, not as a faster federal search box. The workload was state and local: small municipalities, odd counties, and scattered portals that never show up in one place. Across that coverage, the engine screened 6,090 candidate solicitations, cut 60.4% as noise, and returned 2,413 worth a human look. Analysts spent time on fit and pursuit decisions instead of hunting notices portal by portal.
By the numbers
Deloitte in SamSearch
Solicitations screened
6,090
Read by the recommendation engine
Worth a human look
2,413
Surfaced as relevant
Cut before review
60%
Removed as poor fit
Hours returned (est.)
245
Analyst time, platform estimate
From the market to the shortlist
How many solicitations the engine read for Deloitte, how many it removed, and how many made it to capture review.
6,090
Screened
100%
3,677
Cut
60% removed
2,413
Shortlisted
40%
Screening volume by week
Weekly solicitations screened versus the smaller set returned as relevant. From live recommendation activity.
Hours returned uses a platform estimate of 4 minutes per solicitation filtered out. Not a stopwatch measurement for this account.
Capabilities in play
What they run in the platform
Capture
One scored feed across 50,000+ SLED sources plus federal, DIBBS, and GSA eBuy, so a consulting team is not babysitting state boards, county sites, and tiny municipal portals per client.
Manage
Pursuits that clear the filter move into a GovCon pipeline instead of dying in an inbox or a generic CRM stage.
Old way vs. SamSearch
What changed in the pursuit process
| Before | With SamSearch | |
|---|---|---|
| Opportunity discovery | State boards, county sites, and municipal portals checked one at a time, plus whatever federal work still lived on SAM.gov. | One engine across 50,000+ SLED sources and federal markets, reading each solicitation and scoring fit. |
| Bid/no-bid triage | Every county and municipal notice gets a human skim before anyone knows if it matters. | 60.4% cut automatically. Only 2,413 of 6,090 needed a second look. |
| Multi-client capture | One merged feed that mixes every client's geographies, agencies, and NAICS codes. | A distinct AI profile per client, each with its own daily shortlist. |
Platform note
One AI profile per client engagement
A consulting practice is not chasing one company's contracts. It is chasing every client's geographies at once, including the small municipalities and counties that never appear in a single state portal. SamSearch supports a distinct AI profile per engagement, each with its own NAICS codes, agencies, certifications, and daily recommendation pass across 50,000+ SLED sources. Ten clients means ten shortlists, not one unusable firehose.
Client Profile A
Daily recommendations, this profile only
Client Profile B
Daily recommendations, this profile only
Illustrative: two placeholder client profiles under one account, each scored against its own criteria. Not a screenshot of any specific customer's data.
FAQ
What does Deloitte use SamSearch for?
In only 6 weeks, a Deloitte team ran SamSearch as SLED capture automation, not as a faster federal search box. The workload was state and local: small municipalities, odd counties, and scattered portals that never show up in one place. Across that coverage, the engine screened 6,090 candidate solicitations, cut 60.4% as noise, and returned 2,413 worth a human look. Analysts spent time on fit and pursuit decisions instead of hunting notices portal by portal.
What results are documented here?
6,090 SLED and multi-source solicitations screened; 60.4% cut as poor fit before analyst review; 2,413 shortlisted for human capture review.
How were the results verified?
Platform activity comes from SamSearch production data. No award dollars are claimed where a matching public record is unavailable.