Discuss Home · Bootstrapped Podcast · Scribbleton Personal Wiki · HelpSpot Customer Service Software · Thermostat NPS

From name in email communications


Has anyone done A:B testing on what from name works best when sending out emails on behalf of the company?

I’ve seen quite a few variants recently, like:

  • Founder Name
  • Founder from Company
  • Role @ Company

My latest Snappy price-drop update from @ian came from “Ian Landsman” which made me think about this…


Price drop! Those guys must be crazy.

Yeah, I wish I had been doing formal testing on it but I haven’t. I wonder if @patio11 has seen any work on this.


Depends on what you are doing “on behalf of the company”

My transactional receipts to endusers are sent out just with the company name.

My marketing and information to organizations that use my product are in my personal name


No matter you are sending a transactional (personal hello email, password reminder email, etc.) or marketing email (ex: monthly newsletter), try to use the same email address (and name) for maximum deliverability.

On the other hand, we send system notification emails (password reminder emails, weekly report emails, etc.) with “Sendloop” hello{{at}}sendloop.com and all the rest with the sender team member name including company name: “Cem Hurturk, Sendloop” cem.hurturk{{at}}sendloop.com


The kind of folks who are big enough to test emails are generally a few orders of magnitude between the size of business most of us operate at, Ian.

Overwhelmingly, my practice and recommendation is:

  • Name with affiliation for personal communications to customers, lifecycle emails, and drip marketing campaigns. e.g. I think, off the top of my head, that most of my Appointment Reminder emails go out from Patrick (Appointment Reminder).
  • Company alone for messages where both sides desire a totally automatic relationship, like password reset requests and monthly invoices. I think these go out from Appointment Reminder with the address being support@

I generally have most stuff at clients sent under the founder’s signature but some of my clients have one particular person in charge of customer relations and they have her (almost always a her) sign things. One of them I’ll call Laura. Laura is so popular, as a result of being in all the onboarding emails (and being the first point of call for any issue sent into the support system) that customers perceive that Laura is the company.

Laura has received two marriage proposals, one from a gentleman who refers to her as Beautiful Software Lady Laura. This has become an in-joke at the company. Every time they send out another mass email which moves the needle “Beautiful Software Lady Laura strikes again!” echoes in the room.

I suggested taking this and running with it on their plans/pricing page, with Laura’s photo recommending that people buy the middle plan which we wanted to push. They didn’t implement that yet but I so think this would work.


It can still make sense to test other things such as the email subject though, even if you have a smaller list with just 20k subscribers or so (especially as it’s so easy to do with tools like Campaign Monitor). For example, we recently sent a campaign to announce our new TestRail 3.0 release and with just a small variation in the subject line (changing just 2 words, in fact), the winning email had a 30% higher opening rate (with high statistical significance). We were quite surprised that such a small change made such a difference. As the test took us only a couple of minutes to set up, we plan to include an A/B test for all our future campaigns.


Crikey. OK, data wins: definitely do that.

Mind sharing what your two subject lines were? I’m always curious about this sort of thing.


Sure, here are the two subject lines:

Version A: TestRail 3.0 with all-new reporting, API and integrations
Version B: NEW TestRail 3.0: awesome reporting, API & integrations

Here’s the result of the A/B test:

Please note that these numbers are from ~18 hours after sending the two versions and the opening rate of the test would have improved over time if we didn’t select the winner at this point and sent the campaign.