7x Conversions using Automated Imagery

By Ryan Kulp On Sep 10, 2015

2014 was the year of outbound, cold email. But when everyone started doing it, the channel became saturated and less effective. This cheat sheet teaches you new strategies to personalizing more than just words in your email.

In an effort to save you 100+ hours of precious productivity, below are the 5 pillars of cold email mantra:

  • Be personable, use first names vs. “Mr” or “Hi”

  • Don’t email on Monday / Friday

  • Always follow-up, sometimes up to 7 times, to maximize response rate

  • Ask questions, don’t make assumptions about their needs

  • Talk about them, not about you

With that out of the way, let’s talk about cold email v2.0.

Just when the mantra crystallizes, consumer savviness increases in step. As content strategists use data to drive campaign touch points, Gmail launches the tabbed inbox. As savvy marketers find new ways to personalize email (“Hi, I found you on Product Hunt…”), consumers learn what APIs do.

So where do we go from here? Is cold email dead? I don’t think so, and my team set out to prove it.



Step 1: Commit

We created a way to scale image personalization, which you can fine-tune for your business. But you’ll need a few hours and team buy-in to make this process a reality. So before you rush through this guide, try custom imagery with your prospects on a 1:1 basis to see if it works.

Step 2: Setup 

Now that you’re getting higher response rates from embedding custom images in your cold email, it’s time to configure the infrastructure.

Here’s what you’ll need:

  • Google Drive desktop app or Dropbox

  • Cyberduck

  • Photoshop

Dedicated S3 Bucket, preferably with a short and neutral name (this will be client-facing)

Step 3: Automate Image Creation

Since you’ve already made custom images for prospects, it’s time to stop doing that. Open Photoshop and create an Action that automates file creation.

For KISI this was a 3-step process:

  1. Find a prospect’s logo

  2. Resize it to 140 x 40px (Photoshop > File > Scripts > Image Processor > Resize to Fit)

  3. Place logo on mock

Here’s what the KISI Action looks like for Step #3 above:

If you’ll be combining files (KISI combines logos + mockups), be sure to include a ‘save’ and ‘close’ step. Save the file wherever you want, and then File > Close to remove the tab from your Photoshop browser. More on this later.

Now test it. (Note: the asterisks only apply if you’re combining two files, ie: logo + static image).

  1. Open the static image in Photoshop, then select-all and copy the image to your clipboard. Close image.*

  2. Open all the files you want to place on the #1 image. Do this with File > Open > Choose Folder (1 shot).*

  3. Create a folder with your G Drive / Dropbox for the final output, ie: “Placed logos”

  4. Run a batch operation (File > Automate > Batch) using the Action, Opened Files, and Destination > Folder settings. The Folder you choose should be the one you created above in Step #3.

  5. Voilà, you should have a folder full of custom images.

Step 4: Upload Images to S3

We need an S3 Bucket because Amazon doesn’t obfuscate file locations like Dropbox or Google Drive.

Here’s the same image in 3 places:

With Cyberduck installed, connect to your S3 Bucket.

Next, drag your folder of images into your bucket.

Finally, select all the images (cmd+A / ctrl+A) then right-click (ctrl+click) and select Info:

This step is critical. For clients to see the files you upload to S3, you must mark them Public for Everyone. Just click the dropdown cog in the bottom left for Everyone, then the dropdown under ‘Permissions’ to add READ:

If you’re using PersistIQ, create a new template. Click ‘source’ then paste this where you want the image:

<img src="https://BUCKETNAME.s3.amazonaws.com/.png" style=”width:300px"><br>

It should look kinda like this:

You’ll notice a couple things:

  1. We’re using the “snippet” field in PersistIQ to store the file names. Create a consistent naming convention for the files your Photoshop action creates. We prefer using a prospect’s domain, ie: snippet = cocacola for prospect john@cocacola.com.

  2. The 300px width is important for email client readability. Displaying the raw image could not only skew your message text, but it will likely cause a lag in the load-time and thus create a bad user experience.

If you’re not using Persist to send emails, that’s OK too. Just create a merge tag with the file naming convention described above.


One of our clients, KISI, offers keyless entry for commercial buildings with a slick mobile app. And while the product is great, it’s invisible. This is oft a challenge for utilities -- sticky, but not sneezable.

Until you change it. With the image hacking method.

Rather than cold emailing prospects with a tell, KISI emails prospects with a show.

Instead of forcing readers to imagine the solution, we handed it to them on a platter. And we did it at scale, to the tune of 100s of prospects weekly.


This cheat sheet will require approximately 10 hours to execute.


  • Leads, or ability to source leads (minimum 100-250)

  • Photoshop

  • AWS S3 bucket

  • S3 Browser, ie Cyberduck

  • Google Apps

  • Google Drive desktop application


Once you nail this process, adding custom imagery to your cold outbound email should boost response rates and provide serious ROI for the 10-15 minutes it takes to do it.

If you’ve completed the steps above and are now sending cold email 2.0, we'd love to hear about it! 



Trying to grow your business? I can help. Shoot me a message over at my Geek Page.

Click here to contact Ryan



Topics: outbound, cold email

Ryan Kulp

Written by Ryan Kulp

User acquisition expert, leveraging frameworks from 100+ marketing books to help you grow.