Echelon Insights

§1About

✎ Drafted by AI from the public sources cited. Each sentence links to its source.

Echelon Insights describes itself as a strategic research firm that combines traditional polling with digital research, analytics and modeling.^1,2 Its team page says it has worked since 2014 and lists Kristen Soltis Anderson and Patrick Ruffini as founding partners and Kai Chen Yeo as a partner.²

Kind

Commercial firm¹

Founded

2014²

§2Who it works with

Why this matters: Owners, news partners and the firms behind a poll's interviews shape what gets asked, how, and what gets published.

Published with

Names its polls have been published under, alongside other pollsters or news outlets.

• Echelon Insights/GBAO: GBAO · 2 polls listed
• Echelon Insights/Tusk Strategies: Tusk Strategies · 1 poll listed

Who paid for its polls

§3Whom it works for

Why this matters: A firm that also works for one party's campaigns can still publish accurate polls, but readers deserve to know.

Works for Republican candidates or groups. The Echelon Insights homepage shows the National Republican Senatorial Committee and the Club for Growth among the organizations it works with, next to a think tank, a foundation, nonprofits, a trade group and The New York Times, and names no Democratic party committee. Its team page says its clients include electoral campaigns and nonprofits but does not name a party.^1,2

• National Republican Senatorial Committee. The Echelon Insights homepage shows the National Republican Senatorial Committee among the organizations it works with.¹

Of its 145 polls listed since the 2024 election, VoteHub flags 0 as paid for by a Democratic-aligned sponsor and 0 by a Republican-aligned sponsor⁴.

Federal campaigns and committees that paid it

From 2023 through 2026, 1 federal committee reported paying Echelon Insights $5,000 for polling, surveys or research^5,6,7,8. By the party each committee reports to the FEC: 0 Democratic, 0 Republican, and 1 none (most PACs and super PACs report none).

§4How its numbers compare with other polls

Why this matters: A firm whose polls run steadily to one side of everyone else's is worth reading alongside others.

Each of its polls since the 2024 election is compared with other pollsters' polls of the same question whose field period ended within 14 days, each other pollster counted once⁴. The figure is the average gap, with a 95% margin. This compares it with other polls, not with the truth: the others can be the ones that are off.

Party margin

R+0.7 (±0.9) across 33 polls of the generic ballot and 2026 races, in 12 contests.

Net approval: Donald Trump

+1.2 points (±0.9) across 23 polls, against other polls of the same question.

Net favorability: JD Vance

+1.0 points (±1.7) across 20 polls, against other polls of the same question.

Net favorability: Donald Trump

+0.4 points (±1.4) across 18 polls, against other polls of the same question.

D+ means its polls showed Democrats doing better than other polls did; R+, Republicans. A figure from fewer than 5 polls isn't shown.

§5Record against results

Why this matters: The one test that doesn't depend on other polls: how close a firm's final polls came to what voters did.

From FiveThirtyEight's final public data^9,10: general-election polls between a Democrat and a Republican in the last three weeks before each election, compared with the certified result. FiveThirtyEight rated each name polls were published under separately. Across all 8,957 such polls, the average miss was D+0.7, so the relative figure takes out the misses every pollster shared that year.

ABC News closed FiveThirtyEight in March 2025; its data stops at the 2022 elections. 2026 results will be added once states certify them.

§62026 record against the count

Why this matters: How close this pollster's last 2026 polls came to how people voted, measured the way FiveThirtyEight measured earlier elections.

7 polls of 7 races between a Democrat and a Republican, with field periods ending in the 21 days before November 3, 2026, compared with each race's count. Some of those races are still being counted, so these figures use the unofficial count and will change until states certify.

Average miss

R+4.9

Relative to other polls of the same races

R+1.0

Typical size of miss

5.2 pts

Error is the poll's D−R margin minus the count's, as shares of all votes in the race; D+ means it showed Democrats doing better than they did. From 7 polls in all.

§7Its polls

§8What we looked for and didn't find

• Its website does not state where the firm is based or name a chief executive.
• No standing media partner was found; The New York Times logo appears among the organizations it works with, without a description of the work.
• No current sample source was found; the Lucid detail comes from a 2022 survey.
• Searches found no AAPOR statement, court ruling or filing, regulator action, aggregator statement about excluding or reweighting its polls, or retraction by the firm.
• The Silver Bulletin ratings table and the VoteHub methodology pages could not be read, so aggregator checks rest on published text from aggregators and news searches.