Minute by Minute: A New Plyto Customer's First Campaign

A new Plyto customer kicked off their first campaign last week. I sat with them for the full ride: from blank workspace to first booked meeting. Here's what actually happened, including the parts that broke.

The company is a 14-person B2B services firm. They came from a stack of HubSpot Starter, Mailchimp, Google Ads managed by a freelancer, and a Notion doc pretending to be a pipeline. One owner, one marketer, two SDRs. Goal for the first campaign: book qualified discovery calls with operations leaders at logistics companies in the Midwest.

0:00 – Workspace spin-up

We created the workspace and connected Google Ads, Gmail, their domain, and Stripe. Stripe matters because Plyto uses closed revenue to back-propagate attribution; without it you're guessing. The connections took about 11 minutes, most of that spent chasing down the Google Ads admin login the owner hadn't touched since 2024.

First mistake: they connected the marketer's personal Gmail instead of a shared inbox. Replies to the first send would land in one person's inbox with no visibility for the SDRs. We caught it later, but noting it here because almost everyone does this.

0:18 – Ingesting what they already had

We pointed the agent at their existing CRM export (a 3,400-row CSV), their website, and the Notion pipeline. Plyto's enrichment pass took about nine minutes. It flagged 812 duplicates, 290 contacts with bounced emails from prior sends, and 44 accounts that had already closed as customers but were still being emailed as prospects. That last group is the one that quietly kills sender reputation.

The agent asked whether to suppress the closed-customer list from prospecting sends. Yes. Moving on.

0:35 – Defining the campaign in plain English

The marketer typed roughly: *"Book 15 discovery calls this month with VPs or Directors of Operations at logistics companies, 50–500 employees, in IL, IN, OH, MI, WI. Avoid freight brokerages. Our ICP is warehousing and 3PL."*

Plyto proposed a three-channel plan: a cold email sequence to 600 enriched contacts, a LinkedIn-targeted ad set with a $40/day budget, and a retargeting layer for anyone who visited the pricing page. It also drafted the landing page.

What went right: the ICP filter correctly excluded freight brokers, including a few edge cases where the company described itself as "logistics" but was really a broker. The agent cross-referenced website copy, not just industry codes.

What went wrong: the first draft of the cold email opened with "I noticed you're in logistics." Generic. We rejected it and told the agent to pull a specific hook from each company's site or recent news. The second draft referenced actual warehouse locations and, in one test case, a recent facility expansion announcement. Much better. This is the kind of iteration the agent needs a human to steer on turn one.

1:12 – Compliance and sending infrastructure

Plyto warmed a new sending subdomain rather than burning their primary. It also auto-generated the unsubscribe handling and a physical address block (CAN-SPAM) using the business address on file. We had to correct the address — the one on file was an old coworking space. Lesson: check the business profile before the first send.

The LinkedIn ads needed manual creative approval. Plyto drafted three variants; we killed one for being too jargon-heavy and approved the other two.

1:40 – Launch

Emails scheduled to drip over four business days at 60 sends per day from the warmed subdomain. Ads live. Retargeting pixel confirmed firing on the pricing page.

Day 1 results

• 60 emails sent, 41% open rate, 3 replies (2 positive, 1 "not now")
• Ads: 2,100 impressions, 14 clicks, 1 demo request from the landing page
• Attribution: the demo request came from someone who had also opened the cold email that morning. Plyto attributed the lead to email as the first touch and ads as the assist. The old stack would have given 100% credit to the ad.

Day 3: the first real problem

Reply rate dropped on day 3. The agent flagged that the follow-up email in the sequence was landing in Promotions for Gmail recipients. Diagnosis: too many links in the signature plus an image. It auto-proposed a plain-text variant, we approved, and the next batch landed in Primary for the sample it tested.

Also on day 3: an SDR replied to a prospect from his own inbox instead of through Plyto. That reply didn't get logged against the deal, so the follow-up cadence kept firing. The prospect got a sequence email the next morning that read as if the SDR hadn't just talked to them. Awkward. We set a rule to pause sequences when a reply is detected on any connected inbox, and added the SDR inboxes to the workspace.

Day 7 totals

• 420 emails sent across the sequence
• 11 positive replies, 6 booked calls
• $280 in ad spend, 4 landing page conversions, 2 booked calls
• Pipeline value from booked calls: $47,000 projected ACV
• Cost per booked call: $35 blended

Two of the eight booked calls had multi-touch journeys that would have been invisible in their old setup: ad click, no conversion, cold email open two days later, direct visit to pricing page, demo request. Plyto stitched the whole path.

What to steal from this

1. Add every inbox your team uses on day one. Orphaned replies break sequences and burn trust.

2. Reject the first draft of anything generic. The agent gets better with one round of specific feedback; it rarely gets there alone.

3. Suppress existing customers before the first send. Nobody checks this and everybody should.

4. Connect Stripe or your billing system immediately. Without closed revenue, attribution is just click counting.

5. Warm a subdomain. Protect your primary sending reputation like it's a bank account.

First campaigns are messy. The point isn't a flawless launch — it's a loop that gets measurably sharper by week two.