A Funding Company Whose Cold Email Wasn’t Working
In August 2026, our team took over cold email for a B2B funding business that provides funding to small businesses across Canada.
They came to us with a history in the channel. An earlier attempt had drawn some interest but no applications, so the team had set cold email aside. The campaign on record from May went to 1,041 businesses. Nine of them replied, a reply rate of 0.86%, and not one of the nine wanted to hear more.
What Was at Stake
Every lead the company worked came from one channel, and cold email was its attempt at a second. The bar was plain. It had to earn back its cost early on, and it had to produce people who would fill out an application or take a call from the team.
What We Found in the Campaign
We opened the campaign to see why it had not worked. It was one email in two versions, with no follow-up after it. One version had been pasted in with its old formatting still attached, so it was barely readable.
The subject line was “quick question” next to the reader’s first name, which cold email has worn out. The personalization dropped a scraped job title, company name and industry into a template, so nothing in it told the reader the note had been written for them, and the company names went out exactly as they were scraped.
The pitch then opened on “we help business owners,” which told the reader that a company was selling to them. Because the email had no case studies in it, the promise of fast working capital was hard to believe.
The list had problems as well. Of those 1,041 businesses, 29 bounced, a rate of 2.79%, which is a sign the list needed verifying before anyone sent to it.
The old email
Sender name
What We Changed Before Sending Another Email
We left the funding product alone and rewrote everything around it. The company names were cleaned up, and “quick question” came out of the subject line. The email now came from one person, whereas before it was positioned as if it was coming from a company. We added case studies and a follow-up that landed on the same thread. The first email asked for one thing, which was whether the person wanted the details, and anyone who said yes received the application link by reply.
We also rebuilt the sending setup before anything went out. The company’s old mailboxes had been disconnected when its subscription lapsed, so nothing was connected to send from until the account was reactivated and the mailboxes were warmed back up. The emails went out from 30 warmed mailboxes at 20 sends a day each, on weekdays from 7am to 7pm Eastern, with open and click tracking off and sending stopped as soon as someone replied.
We verified the list before sending, and the list itself was businesses that had applied for funding before. Each person got two emails, with the follow-up four days after the first.
The campaign, before and after
What the New Email Looked Like
The new email opens on the one thing we knew about the reader, which is that the business had applied for funding before. It asks why that application did not go through, a question this list could answer and a stranger could not.
It then names the two reasons people usually give and answers them, because approval here is based on the deposits going through the account instead of borrowing history. A welding supplier who secured $50k two days later gives the reader something to believe, and the soft pull and one business day decision take away the reasons to wait. It closes on a single question that only needs a yes.
The follow-up is shorter. It lands four days later on the same thread, acknowledges that the person is busy and offers to send more information.
The new email and its follow-up
The First Day of Sending
The first emails went out at 7am on August 21. By that evening, 547 had gone out and 15 people had replied, a 2.7% reply rate and about three times May’s. The bigger change was in what the replies said, because people were saying yes instead of asking to be taken off the list.
The list is a big reason it worked so fast. The company knew these people had applied for business funding in the past, so the email reached them on a subject they had already raised their hand for.
The Final Numbers
The two campaigns reached 3,522 people and 233 of them replied, a reply rate of 6.62% against 0.86% in May. Of those replies, 121 were interested, up from none in May.
Counting only those 121 interested replies, the response per person reached still comes to about four times May’s total reply rate, and May’s figure includes every reply it got, interested or not.
Reply rate per person reached
What Happened When the Owner Asked for a Link
Within days of launch, the owner asked us to put the application link directly in the first cold email. We advised against it. Inbox filters can read a link in a first email from an unfamiliar address as a sign of promotion or spam, which can hurt deliverability for the whole campaign, and it asks the reader to click before they have agreed to anything. The owner wanted to try it, so we wrote that version too.
It ran in two campaigns and reached 2,666 people. Of those, 101 replied, a reply rate of 3.79%, and 22 were interested. That is still well above May’s 0.86% and well under the 6.62% and 121 interested replies from the campaigns that saved the link for the reply. The link campaigns also differed in the opener, the case study, the follow-up, the mailboxes and the start date, and one of them went to a different list.
Why a Change at the Top of the Funnel Carries Through
Nine replies with no interest became 121 businesses saying yes to the same funding product, and every one of them could be sent the application link. That number sits at the top of the funnel, and each rate below it multiplies from there, so a change at the top carries through every later step. Hold those rates and the entire funnel comes out four times bigger.