Our first MailLift concept and process diagram
Performance • Feedback • Revision
Discussions With Customers
Early on, we started trying to write specs and prototype implementations for MailLift. We stopped when we encountered too many questions we couldn’t answer. Questions like: How do our customers tell us what they want written? Should we offer custom stationery? What if our customers want to use their own stationery? What if our customers want to add the handwritten note to a gift? What if our customers want their letters written today and delivered next month? What if our customers want to sign the letters themselves after we hand write them? Will anyone notice if the letters are postmarked in Austin? What kind of pricing model do we use?
The process of designing software to address all of these contingencies could take weeks… so we stopped thinking about software and started talking to potential customers.
Sales attempts #1
In our first discussions, we hypothesized that customers wanted an API for handwritten letters
, similar to Twilio, where they could write MailLift integrations into their own CRM software. At this time, we were also positioning MailLift as a customer retention play instead of customer acquisition after significant feedback on Quora.
After chatting with multiple potential customers, it became clear that an API for handwritten letters
was nice, but the concept did not resonate with our potential customers (as many of them were non-technical).
We learned that we needed to focus more on our own customer acquisition, instead of the technology, and explain the value in a ‘non-technical’ way.
Sales attempts #2
Our next hypothesis involved trying to sell the concept of software that would make handwritten mail merge easy. With the help of our designer, Adrian, we created a few mockups to show potential customers.
Note: We did not write any code for the mockups
Again, our potential customers understood what we were offering but they wanted something even easier. The potential customers who had the greatest value to us (who we expect can generate $40,000+ per year in revenue) asked if we could integrate direct mail with Salesforce
In a long discussion, we uncovered a new way to frame our conversations with customers and investors that made it easy for everyone to understand where MailLift could create the most value. All without writing a line of code.
Learning With a Customer
We secured a pilot with Granite Medsystems to help them convert sales that were all but lost. Granite ordered 127 letters (sent to 12 big companies they wanted as customers) written on their stationery with one of their sales staff’s cards included.
We worked with Granite in email and even drove to their offices to pick up stationery and supplies. We’ll need to streamline these processes when we grow, however at this point we’re most interested in proving that the concept works.
Over the course of a weekend and with two handwriters, we completed the order and shipped the letters to Granite’s prospective customers. (In that period we also learned a lot about how our handwriters work and how to improve the process.)
As of today, a couple weeks after shipping and checking with Granite MedSystems’s sales staff more than a few times, 6 of the 12 companies that would otherwise have been written off have responded to the letters and set up meetings
to meet the Granite sales person. While the data is not statistically significant, a 50% conversion rate on our first production run is an impressive customer testimonial! (24x the direct mail average
according the Direct Marketing Association)
Learning With Investors
We believe that people with money give the best advice and direction. Whether it’s customers buying or product, or investors purchasing equity. Both parties want to minimize risk, increase return, and have the best possible experience. Where investors stand out, is with their experience. They are direct, to the point and rough.
We set up meetings with over 30 investment groups around the country to accelerate our learning. Our pitch changed OVER AND OVER and we believe it it was for the best. Any time we had a hypothesis, we’d through it in the deck. No one is quicker to point out flaws than the people who’ve been there and done it.
We’ve documented (audio/video/transcript) and learned from over 25 hours of conversations between MailLift and some of the brightest investors; Hyde Park Angels, Twin City Angels, Central Texas Angel Network, Harvard Business School Angels, TechStars, 500 Startups, WAFFT, Wisconsin Investment Partners, Angels on the Water, etc., etc.
Operations and Processes
Handwriting letters is labor-intensive. If a customer orders 10,000 letters, what do we do to fill that order?
This question sparked a number of other questions we started investigating.
How do we find handwriters? How much do we pay them? How do they work? When and where do they work? How many letters can a single handwriter produce in a given time period?
How do we find handwriters? What do we pay them?
Our first experiment involved sitting in an area with heavy college student foot traffic and asking random passers-by to hand write samples. We really enjoyed this.
We started offering around $20-25/hr to get people interested and lowered the offering price down to $15/hr with no change in interest.
College students and their parents flooded in to show us their skill and we ended up running out of the tiny amount of paper we brought with us.
This experiment showed us three things:
• Handwriters are open to working for $15/hr part time
• Finding people with legible handwriting is simple if you get in front of enough people
• We need a quality control process to catch and correct errors
We’re always watching our margins so it made sense to ask: Can we still adjust what we pay our handwriters?
By this time, we had a few early pilot stage customers who were willing to explore MailLift’s value-add in their business. Within the first week of opening our sales pipeline, we had orders from Michael Best (huge law firm) and Granite Medsystems (medical manufacturer). So began the race to deliver on our product.
We posted an employment advertisement for handwriters at $10/hr part time and received about 15 promising responses in the first 24 hours! We kept the advertisement up for a few weeks and it continued to generate promising applications! Handwriters are still interested in working for $10/hr. part time, not $15/hr. At this point, we have not found a lower limit to our labor costs. $10/hr is a number we are comfortable with, so we moved on.
We interviewed our first 8 handwriters in a large group. Many of these handwriters had other jobs and were looking to supplement their income… preferably by working from home. We long suspected that it made sense to have a distributed network of writers and it was good to know that our handwriters were on the same page. We plan to design our software to work on iPads for people working from home.
Why pay hourly?
MailLift is still perfecting the process and how we organize operations. It’s constant performance, feedback, and revisions. We’re optimizing everything from how our handwriters improve quality and avoid mistakes, to how our software UI scales to handle thousands of pieces of mail. We want good relationships with our handwriters. Many of whom, are well educated and have useful suggestions about how we can help them improve speed and quality. To that end, we’re hiring handwriters as part time hourly staff to work with us through the first few major orders and help us make the best possible processes and software.
Investors and mentors alike have suggested that we move to per-unit-produced compensation for our handwriters
. One potential investor has experience with highly skilled contract labor and offered insights on how the per-unit-produced approach better aligns our interests with our writers’. We fully intend to test these insights in our business when it makes sense.
The handwriting process
At the beginning, we defined our v1 process with a simple whiteboard sketch.
We continually tried to reduce the number of steps required to ship a high quality product but our initial process sketch is the standard for how we operate today with one exception: we photograph every letter before we stuff, seal, stamp, and send.
Example letter with the spec sheet. We are designing our software to replace the letter spec sheet.
How many letters does a handwriter produce in a given amount of time?
When interviewing handwriters, we had each handwriter write copy from one of our sample letters for one minute. We counted each handwriter’s Words Per Minute to get an idea of how many letters they could produce. This first group of handwriters varied widely from 9 WPM to 32 WPM with print. We also collected cursive samples but many of our writers’ cursive samples were lacking in quality compared to print.
In general, the handwriters with the best penmanship are slower, but there are a few exceptions to that rule. This is further validation that we should focus on a way to pay our handwriters based on the number of letters they produce instead of hourly.
After a few handwriting sessions, one of our favorite handwriters produces about 3-4 letters an hour. This is more expensive than we would prefer but his command of penmanship is so excellent that we’re willing to work with him to improve his speed.
We’re also using MailLift as a marketing tool -- reaching out to prospective customers and investors. Reactions are largely positive. Generally speaking, people love receiving handwritten letters.
Big thank you’s to everyone that provided feedback:
And the groups of people with:
Hyde Park Angels
Hyde Park Ventures
New World Ventures
Corner Stone Angels
Space Center Ventures
Twin Cities Angles
New Capital Fund
Angels on the Water