Consultancies vs. regular companies: who actually responds?

Every Apollo search used the same diluted filter, so each campaign emailed a mix of consultancies/auditors and regular operating companies ("preparers"). Splitting the results by who the recipient actually is: consultancies out-perform at every funnel stage, roughly 2–4× — but the two twists at the bottom complicate the obvious conclusion. Data: Overloop era Apr–Jun 2026 (send base, n=2,941) and July engaged funnel (n=759), reports of 2026-07-25.

≈2×Consultancies vs. regular companies on open, click, and reply rate (true send base, n=2,941)
4.0×More likely to become a top-3-tier lead once engaged (11.5% vs. 2.9%, July funnel)
29% → 1.7%Verifier/auditor lists: click rate collapses to reply rate — huge curiosity, no conversations
1 dealThe only active deal (Shadowbox Studios) is a regular company, not a consultancy

The segments

Consultancies
firms paid to do sustainability/ESG work for clients: consultants, advisors, auditors, verifiers, assurance providers (e.g. KlimAktiv, Verdani, EcoEngineers). They would use Sustentra on client engagements.
Regular companies ("preparers")
operating companies reporting their own emissions data: manufacturers, real estate, food, apparel (e.g. Shadowbox Studios, Adafruit, Kyocera AVX). They would buy Sustentra for their own compliance.
Software / carbon-tech
vendors of sustainability, LCA, or carbon software and carbon-market platforms (e.g. iPoint, Ecochain, GLYNT, Proba) — competitors or potential partners, not typical buyers.
Public sector
government bodies, utilities, nonprofits, chambers, foundations, and trade forums (e.g. Northern Water, Minnesota Chamber, USISPF).
Investors
VC, PE, and asset managers with ESG mandates (e.g. HTGF, Quantum Capital, Green Star Royalties) — interest is usually portfolio-level reporting.

Assignment: hand-labeled for the 48 most-engaged leads; keyword heuristic on job title, company name, and email domain for the rest (see footer for details).

1 · Top of funnel — real send denominators (Overloop era)

2,941 contacted with known outcomes. Consultancies opened at 12.6% vs. 6.9% for regular companies, clicked at 7.2% vs. 3.2%, and replied at 1.8% vs. 0.9% — a consistent ~2× advantage at every stage.

SegmentContactedOpenedClickedReplied
Consultancies27812.6%7.2%1.8%
Regular companies2,3626.9%3.2%0.9%
Public sector3008.0%3.3%0.7%

Classified by job title and email domain (consultant/auditor/verifier/assurance terms → consultancy); one investor contact omitted (n=1). Hand labels override the heuristic for the 48 deeply-engaged leads.

2 · The verifier experiment — reachable, curious, silent

The Overloop era included hand-curated "US carbon verifier" and "US auditor" lists (354 people). They opened at 38% and clicked at 29% — an order of magnitude above the generic Apollo-import lists — yet replied at just 1.7%, the same order as everyone else. The verifier audience is demonstrably reachable and curious; the funnel loses them between the click and the conversation.

Caveat: part of the open/click gap is likely deliverability — the curated lists were small and clean, the "all other lists" bucket (n=2,587) includes a 1,900-contact bulk Apollo import. The click→reply collapse within the verifier lists is unaffected by that confound.

3 · Depth of engagement — July funnel (759 leads with any activity)

Among people who engaged at all in July, consultancies go ~3–4× deeper than regular companies on every metric: repeat sessions, reaching a ranked tier, replying. Regular companies supply the volume (475 of 759) but almost never go deep. The small software/carbon-tech bucket goes deepest of all — those are peers (iPoint, Ecochain, GLYNT, Proba) scoping a new entrant, not buyers.

SegmentEngagedTraced to site2+ sessionsHigh-intent actionTop-3 tierRepliedConfirmed +Confirmed −
Consultancies13951.1%6.5%0.7%11.5%4.3%12
Regular companies47546.3%1.9%0.2%2.9%1.3%10
Software / carbon-tech2356.5%8.7%8.7%26.1%13.0%00
Public sector7352.1%5.5%0.0%9.6%4.1%10
Investors4862.5%2.1%0.0%8.3%6.2%00

Rates are shares of the engaged base (at least one open/click), not of everyone emailed — July send denominators per segment aren't recoverable because every campaign mixed audiences. "High-intent action" = demo click, ROI calculator, early-access request, partner inquiry, or booking. "Confirmed +/−" = replies whose actual text was verified (17 reply texts remain unverified in the connect@ mailbox).

4 · The two twists

Twist 1 — revenue points the other way

The only active deal is a regular company

Shadowbox Studios (film-studio infrastructure, a preparer with its own BAFTA-Albert reporting obligation) is the one lead with wallet-shaped signals: meeting held, NDA signed, sample invoices and questionnaire in flight. Meanwhile, two of the three confirmed "no" replies came from consultancies — including EcoEngineers, who declined after a full demo ("sticking with our current audit procedures"). Consultant curiosity may be professional interest, not buying intent.

Twist 2 — the deepest engagers aren't customers

Peers scope harder than prospects

The software/carbon-tech bucket tops every depth metric (26% reach a ranked tier, 13% reply) — but it's iPoint, Ecochain, GLYNT and Proba: competitors and potential partners studying a new entrant. High engagement from this bucket inflates the numbers without representing pipeline. The same logic applies, in softer form, to consultancies evaluating the competitive landscape.

What this means for the calls

The ICP question, sharpened

Not "which industry?" but: is the buyer a consultancy that would run Sustentra on client engagements, or a mid-size preparer with a compliance deadline? The engagement data leans consultancy; the only revenue evidence is a preparer. This week's calls should test both hypotheses — the segment-guess column on the call sheet marks which experiment each call runs.

Sources & caveats
  • Send base: contacted-leads.csv (Overloop, Apr–Jun 2026, 2,941 contacts with send/open/click/reply counts).
  • Engaged funnel: growth-agent/data/leads_ranked.json (759 leads, Snov + PostHog cross-referenced, report of 2026-07-25).
  • Segment labels: hand-classified for the 48 top-tier leads (data/segment_guesses.json), keyword heuristic on company/title/domain for the rest (segment_performance.py).
  • Reply sentiment: data/reply_sentiment.json — 17 of 29 reply texts still unverified pending the connect@sustentra.com mailbox export.
  • All four Apollo filter versions shared the same job-title list, so no campaign was a clean single-audience test; per-lead classification is used instead.

Europe clicks, America converts

The July campaigns split cleanly by market: localized German, Dutch, and French sends vs. English/North-American sends. The localized EU campaigns earned double the click rates — and a third of the conversion. Interest is not the problem in Europe; the step after the click is. Data: metrics.json send base (2026-06-19→07-19), July engaged funnel (n=759), Short.io human-click validation.

67% vs 33%Click rate, July Germany vs. the best EN campaign — localized EU emails get twice the clicks
715 → 3Human clicks on "book a meeting" links → meetings actually booked. All 3 came from North America
2.8×North-American engaged leads reach a ranked tier at 9.2% vs. 3.3% for EU engaged leads
481 → 0EU booking-page clicks (DE+NL+FR) that produced zero meetings — the boss's funnel leak, confirmed

1 · Send-base rates by market

Localized EU campaigns dominate opens and clicks; replies flatten everyone to under 1.5%. Netherlands is the only market where clicks turned into a meaningful reply rate.

CampaignMarketSentOpenedClickedRepliesBounce
July Germany 300DE29777.8%67.3%13.0%
July Netherlands 150NL14877.0%69.6%22.7%
July France 150FR14571.7%54.5%03.4%
Overloop Unsent (EN core)US/EN78964.4%33.2%33.5%
Overloop FollowupUS/EN16637.3%5.4%00.6%
Warm OutreachUS/EN2516.0%0.0%00.0%

Open/click rates are raw Snov tracking and include some email-security scanner clicks (225 scanner clicks were identified and rejected in the lead-matching step); the cross-market comparison is still valid since scanners hit all campaigns. Audience mix differs too: the EN core list skews consultancy-heavy (see tab 1), the EU lists skew preparer-heavy — language and audience are confounded.

2 · The booking-link leak, by market

Every campaign carried a "book a meeting" link to the same Outlook booking page. Humans clicked it 715 times. Three meetings were ever booked through outreach — all with North-American prospects. The German link alone took 243 human clicks and produced zero.

Human clicks per Short.io link (scanner clicks already removed). The booking page is an English-language Outlook "Book with me" page with US-timezone defaults — the most likely break point for EU visitors who just read a fully localized email. Homepage links leaked similarly: 598 EU human clicks on localized-campaign homepage links, 1 EU meeting ever booked (via reply, not the page).

3 · Depth once engaged — EU vs. North America

Region (by campaign)EngagedTraced to site2+ sessionsHigh-intent actionTop-3 tierRepliedConfirmed +
North America / EN campaigns39330.5%4.8%0.8%9.2%4.1%2
Europe (DE/NL/FR campaigns)36668.9%1.6%0.3%3.3%1.6%1

The inversion is the story: EU recipients click through to the site far more often (68.9% traced vs. 30.5%) but then bounce — only 1.6% ever return for a second session, vs. 4.8% of North Americans. Wide curiosity, shallow commitment — or a site that doesn't speak their language once they arrive.

4 · What this means

Twist — clicks lie in Europe

EU click rates are a vanity metric right now

Localized subject lines earn the click, then everything after the click is English: the site, the booking page, the calendar. EU visitors arrive at 2× the rate and convert at ⅓. Fixing the post-click experience (localized landing page, EU-timezone booking, German-language follow-up) is worth more than any new EU list.

The exception proves it

The one EU positive replied instead of booking

Thomas Prinz (KlimAktiv, Germany) never touched the booking page — he replied to the email in German asking for company references. The EU conversion path that works today is the reply, not the booking link. vivi's EU calling window (8–11am ET) exists precisely because these leads need a human bridge.

Caveat

Language and audience are tangled

The EN campaigns mailed the consultancy-heavy Overloop list; the EU campaigns mailed fresh preparer-heavy Apollo lists. Some of the depth gap is audience (tab 1 shows preparers go less deep everywhere), not geography. The booking-page leak, however, is unambiguous.

Sources: growth-agent/metrics.json (per-campaign send funnels, window 2026-06-19→07-19); data/leads_ranked.json (engaged depth); lead_intelligence.md § Campaign volume validation (Short.io human-click counts); booked meetings from calendar cross-reference.

One boring list did the work

Seven campaigns, very different jobs. The unglamorous English core list ("Overloop Unsent") produced 20 of the 48 ranked leads — more than all three localized EU campaigns combined — while France, with a 55% click rate, produced exactly zero. Data: metrics.json send base (2026-06-19→07-19), July engaged funnel, ranked tiers of 2026-07-25.

20 / 48Share of all ranked (top-3-tier) leads produced by Overloop Unsent alone
3.4%Netherlands: best sent→ranked-lead conversion (5 from 148 sent) — small but the most efficient list
0Ranked leads from July France 150, despite a 54.5% click rate
9 repliesFrom "LinkedIn for Email Openers" — whose sends Snov never tracked, and whose reply texts are still unverified

1 · Send-base performance per campaign

CampaignSentOpenedClickedRepliesBounceRanked leadsRanked per 100 sent
Overloop Unsent (EN core)78964.4%33.2%33.5%202.5
July Germany 30029777.8%67.3%13.0%72.4
July Netherlands 15014877.0%69.6%22.7%53.4
July France 15014571.7%54.5%03.4%00.0
Overloop Followup16637.3%5.4%00.6%00.0
Warm Outreach2516.0%0.0%00.0%00.0
LinkedIn for Email Openersn/t9+9

"Ranked leads" = High + Medium-High + Medium-Low tier (the 48 on the call sheet; the remaining 4 came from pre-Snov Overloop and small residual sends). "n/t" = sends not tracked: the LinkedIn-for-Openers emails went out outside Snov's tracking window, so its rates cannot be computed — its 9 replies are real events but every one of its reply texts sits unverified in the connect@ mailbox.

2 · Ranked leads produced — the only metric that matters

Absolute counts, not rates. Gray = the LinkedIn-for-Openers leads, which are all unverified-text repliers — they may deflate to near zero once the connect@ mailbox is read.

3 · What this means

Twist — click rate ≠ campaign quality

France: the perfect vanity campaign

71.7% opened, 54.5% clicked, 82% of its engaged recipients traced onto the site — and zero replies, zero ranked leads, zero meetings. High deliverability and curiosity with nothing underneath usually means the offer didn't land for that market, not that the list was bad. Whatever the French emails promised, the site didn't deliver it in French.

Where the pipeline actually came from

The EN core list + replies, not clicks

Overloop Unsent produced 20 ranked leads including the only active deal (Shadowbox) — on the lowest click rate of any live campaign. Its recipients clicked less but did more: replied, returned, used the ROI calculator. When preparing the next two campaigns, clone this list's shape (US, consultancy/preparer mix, 1–200 employees), not the EU campaigns' click rates.

Follow-ups underperform

Overloop Followup: 166 sent, nothing back

The follow-up to non-responders got a third of the core list's opens and one-sixth the clicks, with zero replies or ranked leads. A second identical touch to a cold list is a spent signal — the retry budget is better spent on vivi's calls to the 40 warm names than on another send to the cold 2,900.

Sources: growth-agent/metrics.json; data/leads_ranked.json; csv_exports/08_top3_tiers_combined.csv. Reply-text verification status: data/reply_sentiment.json.

The doers respond; the C-suite doesn't

Splitting the 759 engaged leads by the seniority in their job title: engagement is inversely related to rank. C-level recipients almost never reach the site (11.9% traced); managers and specialists do the actual research; VPs and directors convert best into ranked leads and replies. Data: July engaged funnel (n=759), titles from Snov/Apollo records.

11.9%C-level engaged leads who ever reached the site — vs. 59–62% of managers and specialists
4.0%VP / Director / Head reply rate once engaged — the best of any level
16 / 48Ranked leads who are VP / Director / Head — the largest block on the call sheet
0Confirmed-positive replies from C-level contacts. Both deal-shaped positives came from managers

1 · Depth once engaged, by seniority

LevelEngagedTraced to site2+ sessionsTop-3 tierRepliedRanked leadsConfirmed +
C-level / owner8411.9%3.6%6.0%2.4%50
VP / Director / Head22545.3%3.1%7.1%4.0%161
Manager / Lead31259.3%2.6%4.5%1.6%141
Specialist / IC8161.7%6.2%8.6%3.7%71
Other / unknown title5744.0%3.5%10.5%5.3%60

Levels parsed from job titles (English + German patterns). "Other / unknown" merges 31 unmatched titles and 26 blank ones. Titles come from Apollo/Snov records and can be stale.

2 · What this means

Twist — access is inverted

The higher the title, the colder the click

C-levels open email on a phone between meetings; 88% of the engaged ones never reached the site. Managers and specialists are the ones who actually evaluate tools — 60% visit, and specialists return most often (6.2%). Cold email is a bottom-up channel here: land with the doer, let them carry it up. Both deal-shaped positives (Seth Hauer, Thomas Prinz) are exactly that profile.

Tie-back to tab 1

The filter excluded the one exec who does buy

The v1 Apollo filter excluded owners, founders, and partners by seniority — but at 1–200-person consultancies (the best-performing segment), the partner is the buyer, working hands-on. The C-levels we did reach were mostly corporate executives who delegate. If a future search re-adds anyone, it's founder/partner at small firms — not more corporate C-suite.

For the call sheet

Call the VP/Director block first

VPs/Directors/Heads are the call sheet's largest block (16 of 48) and its best repliers (4.0%). They combine enough seniority to say yes to a meeting with enough proximity to the work to care. On calls with managers/specialists, the goal is the internal champion; with the rare C-level, the goal is the referral downward.

Sources: data/leads_ranked.json (titles + activity); tier counts from csv_exports/08_top3_tiers_combined.csv; confirmed positives from data/reply_sentiment.json.