Monday 13 July to Sunday 19 July 2026. The week the personalised report engine ran at pace: five AI HTS and Tariff Exposure Scans were built and published for named prospects, LinkedIn returned its best acceptance rate of the whole engagement at 39 percent, and twelve replies came back across the two channels. Email volume was deliberately held while the follow-up sequences finished landing on the June and early-July lists.
Published late. This report covers 13 to 19 July and should have gone up on 20 July. It is being published on 3 August alongside the Week 4 report, so the record on this portal is complete. Every number in it was pulled live from the sending platforms on 3 August, not written from memory.
Week 2 ended with five prospects who had said yes to a personalised scan and none of them on a calendar. Week 3 was spent clearing that queue. Five scans were researched, built and published for named companies, each one a real report on that company's own import profile rather than a template with a logo swapped in.
Email volume was intentionally flat. The 1,915 emails that went out were mostly follow-up steps landing on people already contacted, with only 355 new people reached. That was the right call at the time: the June and early-July lists were still working through their sequences, and adding fresh volume on top would have muddied the read on whether those lists were any good.
LinkedIn had its strongest week of the engagement. 69 connection requests produced 27 acceptances, a 39 percent rate, against a 26 percent average across all time. Six replies came back on that channel alone, and four prospects were tagged as interested.
Each of these is a real, hosted report on that company's own customs exposure, built from public import records and the live tariff schedule. They are still live and you can open any of them right now.
| Company | Published | Report |
|---|---|---|
| ARC Global Logistics, Las Vegas | Tue 14 July | Open the scan ↗ |
| Concordia International | Tue 14 July | Open the scan ↗ |
| Packair Airfreight, Fountain Valley CA | Tue 14 July | Open the scan ↗ |
| Interport Logistics, Los Angeles | Tue 14 July | Open the scan ↗ |
| Welke Customs Brokers, Chicago | Fri 17 July | Open the scan ↗ |
All five went to people who had already asked for one, so none of these was a cold send. They are the asset that turns a warm reply into a reason to take a call.
Open tracking is switched off on all your campaigns because open-tracking pixels hurt deliverability, so there is no open rate here and there will not be while we send this way.
| Metric | Result | Read |
|---|---|---|
| Emails sent | 1,915 | Mostly follow-up steps on people already contacted |
| New people reached | 355 | Deliberately low while the existing sequences finished |
| Replies received | 6 | 1.69% of new people reached, the best rate of any week so far |
| Bounces | 12 | On the older June lists, which is where the bounce problem sat |
| Metric | Result | Read |
|---|---|---|
| Connection requests sent | 69 | Across both accounts |
| Connections accepted | 27 | 39%, the highest acceptance rate of the engagement |
| Messages sent | 32 | To people who had accepted |
| Replies | 6 | Matching email's reply count off a fraction of the volume |
| Tagged interested | 4 | Prospects who engaged positively enough to work |
12 replies across both channels off 355 new email contacts and 69 connection requests, plus five personalised reports delivered to people who had already put their hand up. Low volume, high touch, and the highest quality-per-send week of the engagement.
69 connection requests produced as many replies as 355 cold emails. The channel is slower to scale and capped by daily limits, but the response per touch is an order of magnitude better. That is the argument for keeping both channels running rather than pouring everything into email volume.
Five reports delivered in one week, all to people who asked. Nobody has yet said no to receiving one. The offer is not the problem in this engagement and never has been. The problem is the step after it.
Twelve bounces off only 355 new contacts is a rate that would eventually threaten the sending domains. Every one of them came from the older June-sourced lists rather than anything built recently. That finding is what pushed the list rebuild that happened the following week.
Two things came out of this week that shaped everything after it. The bounce concentration proved the June lists needed replacing rather than patching. And the pattern in who was replying, forwarders and brokers with names and titles we could see, became the seed data for the lookalike list built the following week.
That work is covered in the Week 4 report, and the result of it is in the Week 5 report, which produced 32 replies and the first booked call.