Maxionlabs
ClearSpider Outbound: Results and Learnings
A straight account of the 90-day cold-email engagement. What we tried, why each approach did not convert, and what we believe would work better. Built to be useful raw material for whatever ClearSpider runs next.
Over about 90 days we ran roughly 14 campaigns across six strategic shifts, sent close to 40,000 emails, and tested more than 15 distinct angles and offers. It did not produce booked meetings. This document is the honest version of why, because the reasoning is more valuable to you than the result.
2.1%
reply rate (incl. OOO)
The one-line read
ClearSpider sells into a market that is problem-aware, largely already-tooled, and slow to buy, and the pain lives inside each prospect's own inventory system where a cold email cannot see it or prove it. Cold outreach can reach that market and land clean in the inbox, which we confirmed, but without a low-friction way for a not-yet-in-pain buyer to raise their hand, the channel caps near zero almost regardless of how good the copy or the targeting is. The gap was market and offer architecture, not the channel and not the delivery. The fix, which we get to at the end, is a front-end offer.
What we tried, phase by phase
Each phase below is a different bet on where the problem was. We moved on only after a phase gave a clear read.
Phase 1
Contractor angle bake-off
Who: around 2,320 US and Canadian electrical and construction contractors, 100-plus employees, multi-state.
Hypothesis: the problem is real for contractors, we just need the right angle, so find the frame that makes an operations leader respond and then double down on it.
What we tested: four angles run side by side (an insider question, a pre-delivered scaling-math benchmark, an anti-Procore objection teardown, a hiring-signal trigger), plus a few one-to-one personalized pieces.
Result: ~2,320 sent, 2 replies, 1 soft positive, 0 meetings.
Why it did not work: the miss was upstream of the angle. Many contractors who replied said their subcontractors handle materials, so they carry no inventory and have no need. Firmographically they fit, but the pain was not there. No amount of angle-tuning fixes a targeting miss.
Learning: the list matters more than the angle. If the list does not feel the pain urgently enough, they will not care, and no angle will change that.
Phase 2
In-market contractors, the switch playbook
Who: 2,245 contractors running a detectable, failing incumbent system (Viewpoint Vista, Sage, Foundation, Procore, and similar), tiered by how clearly the signal named the system.
Hypothesis: clone the situation of the one account that did convert cold, a contractor who already had a system that was not working and was actively evaluating alternatives, and catch others in that same replacement moment.
What we tested: a migration-playbook offer aimed at prospects whose job postings named a dying inventory or ERP system.
Result: 2,245 sent, 0 replies, 0 positive, 0 meetings. The looser tier also bounced at 5.4%, the worst list decay of the engagement.
Why it did not work: a migration offer only speaks to the small share already shopping for a replacement, and signal-based targeting narrows the pool to exactly that sliver. You end up with a tiny list and an offer that cannot create urgency, only catch it.
Learning: in-market signals cap you at the already-shopping few. They cannot manufacture demand, and the narrower you signal-target, the smaller and more fragile the list becomes.
Phase 3
Pivot to distributors, one engineered sequence
Who: around 1,800 contacts at 447 tightly scored multi-branch industrial, electronic, lab, and plumbing distributors, 200-plus employees, seeded on lookalikes of your best accounts.
Hypothesis: the vertical was wrong, not the channel. Distributors, not contractors, are the real ICP, and a carefully engineered sequence on a clean list converts where contractor copy failed.
What we tested: a single three-step sequence using each prospect's own branch footprint as proof, the "what is on the screen is not what is on the shelf" premise, and the Grainger build-versus-buy story.
Result: ~1,800 sent, 2 replies, 1 soft positive, 0 meetings.
Why it did not work: the distributor pivot was directionally right, the numbers ticked up slightly, but we had iterated ten variations of copy on one unproven offer premise. The offer itself was never put to the test alongside alternatives, and no reply-to-meeting motion had been built.
Learning: test offers, not copy. Craft cannot rescue a single unproven offer, and this is what pointed us to the next phase.
Phase 4
Offer-discovery: six offers on stable copy
Who: a broad pool of around 8,970 verified distributor-lookalike contacts.
Hypothesis: the offer is the biggest lever, so hold the copy stable and test six genuinely different offers at once and let the market vote on which pain is sharp enough to act on.
What we tested: six distinct offers, each built on a different pain, stockout risk, winning back lost accounts, freeing trapped capital, scaling to new locations, freight margin, and the Grainger VMI program for larger organizations.
Result: ~8,970 contacts, the largest single block of the engagement, 0 positives across all six offers.
Why it did not work: this is the phase that told us the most. When six different offers on the same market all return zero, the problem is not the wrapper, it is the pain. The mid-to-large distributor pool we reached is already handling this, with rep-managed programs, incumbent ERPs, and spreadsheets that are good enough. Prospects said it plainly in replies: they manage it with reps on site, they already run a materials tool, there is no budget this year. A generic visibility or cash or stockout pitch does not make an already-tooled buyer feel singled out.
Learning: when every distinct offer on a pain returns zero, the pain is falsified for cold, not the offers. This is the single most important finding of the engagement.
Phase 5
Broad volume, a paid incentive, and the replacement-moment rewrite
Who: multi-vertical distributors at higher volume (building materials, MRO, electrical, plumbing, mixed, food), plus a second run at in-market contractors.
Hypothesis: go broad and plain at higher volume, then lift the qualified-meeting rate two ways, a 500 dollar charity donation for a qualified meeting, and copy rewritten into the real language buyers use at the moment they replace a system (old scanners, paper, the last tool nobody adopted).
What we tested: the charity incentive on around 843 reconciled distributors, four replacement-situation copy variants, and several tightly targeted in-market contractor arms.
Result: roughly 2,600 additional sends, 0 positives, 0 meetings.
Why it did not work: an incentive amplifies whatever the pain line selects for, so on a pain the market does not feel, it buys nothing. Volume and money could not out-run a market and offer mismatch.
Learning: you cannot out-volume or out-incentivize a fit gap. An incentive needs a pain that at least one slice already self-identifies with before it can do any work.
The pattern underneath all five
Every phase, whatever we changed, ran into the same wall. Pulling it together:
- The market is problem-aware and already-tooled. Most prospects have solved this well enough with reps, an ERP, or spreadsheets. Nobody is sitting in acute, unsolved pain waiting to be named.
- The pain is private. The gap lives inside each prospect's inventory system. We cannot see it from the outside, so we cannot show a prospect their own leak the way you can with, say, a website or an ad account. That removes the strongest cold-email move.
- The purchase is slow and committee-owned. A three to six month, multi-stakeholder decision is a long way from a single cold email, so even a perfect message is asking for a very large first step.
- There was no low-friction way in. Every campaign asked, in effect, for a demo. For a buyer who is not in pain today, that is the highest-friction thing you can request, so only the tiny in-market sliver ever raised a hand.
Those four are structural. They are properties of the market and the product motion, not of the copy. That is why more copy, more angles, more volume, and even a cash incentive all landed in the same place.
What we believe would work better
The engagement ended with a clear view of the one thing that was missing the whole time.
A front-end offer (the primary fix)
The single change most likely to break the pattern is a front-end offer, sometimes called an introfer. Instead of asking a stranger to book a demo, you give them a real result first, low commitment, and earn the conversation with buyers who are not in pain today but will be in six months, which is most of your market. It answers all four structural problems at once. It lowers the ask, it does not require the buyer to already feel the pain, it does not need you to see inside their system, and it starts the relationship long before the slow purchase decision. There is a simple test for whether an offer will pull: does it give a big result, is it believable, is it fast, and is it low effort for them.
Two shapes fit ClearSpider. The first is done-for-you, where you produce the result and the call is where you build trust and take it further. For example a free VMI pilot on one of their customers, so they watch that account grow and stick before committing, the way Grainger scaled, or a cross-branch leak map built from their own inventory export. The second is a packaged intro offer that solves one problem fast with a number and a deadline. For example putting one customer on VMI in 21 days and lifting that account's reorder volume, or getting a worst-performing branch under 2 percent stockouts in 30 days, with the risk reversed. We would lead with the free VMI pilot on one account, because it fits how distributors grow and it is the stickiest way in. The separate offer directions document lays these out in full.
A separate note we would put in front of you: you have paid for search traffic for close to three years and your site catches almost none of it, because there is nothing between reading the site and a phone-gated demo. A front-end offer plus gating the content you already publish would plug that leak with no new ad spend. It is the highest-return move available to you and it does not depend on cold email at all.
A narrow give-first motion (the complementary cold play)
If cold email stays in the mix, the only version we would run is a narrow give-first play aimed at the small slice showing a dated, live buying signal (a newly advertised program, a fresh hire whose posting names the incumbent, a new branch). To that slice, lead with a genuinely useful free gift rather than a pitch. This was the direction we had moved to by the end. It is slower to stand up and it reaches a deliberately small pool, but it is the one cold motion that respects how this market actually buys.
The honest bottom line
Cold email can reach this market and test every angle, offer, and incentive we could think of, and it still caps near zero, because the pain is not publicly observable and not sharp enough for a cold reader to self-identify. The work that remains is not better copy. It is a front-end offer that lets a buyer step in before they are ready to buy. Build that, and the same channel that struggled here has something real to carry.