Why is it that nobody ever writes much about Customer Discovery failures? Y’know – instances where that first step of Customer Development outright fails. And, even when somebody does write about it, they gloss over the details. As typically told:
- Entrepreneur has idea.
- Entrepreneur finds people to interview.
- Entrepreneur conducts interviews.
- Success ensues.
Even if step #4 doesn’t happen, they still gloss over steps #2 and #3 and skip straight to the findings. And so, here’s a case study of my most recently failed attempt at Customer Discovery.
At some point while growing up, I got it in my head (probably from my Mom) that responsible adults reconcile their bank and credit card statements against their receipts. Meaning, when you get a credit card statement each month, you match each listed transaction against one of the many receipts you’ve kept off to the side for just such purpose. Me being a responsible adult, this is something I did.
I wouldn’t say it was fun, though. The best word I can think of to describe it is “tedious”. Picking a receipt out of the pile and manually (with my eyeballs) scanning (for me) two or three transaction statements each month was laborious and time-consuming.
At some point a year or two ago, I stopped and thought to myself that there had to be a better way. And there was. Me being a software developer, I built a web-based app (website) for myself that sped up this process – reducing the time and tedium by, oh, I’d say 90%.
I was pleased. Eventually, I started to wonder if anybody else in the world would find my creation useful. Useful enough to pay for it. And, if so, whom?
Very quickly, I dismissed the consumer market, for several reasons:
- Consumers are known to NOT value their time, at least not when it comes to rationalizing the purchase of time-saving software.
- I spoke informally with several friends about the tool, and very few (all but one?) did any sort of transaction reconciliation.
- Ostensibly, the reason for reconciling transactions at all is to detect fraudulent and erroneous charges (e.g., purchases made against your stolen credit card number or a waiter giving themselves a bigger tip than you felt they deserved). And this is what I, personally, would sometimes find. But, truth be told, the dollar amount I would recover rarely amounted to more than $9.99 each month (the most I could imagine charging consumers for use of my envisioned website).
So, I turned my attention to the business market. After speaking with just a couple small business owner friends of mine, I shifted my focus to bookkeepers as a possible market (as my one friend said “that’s why I pay a bookkeeper”). However, I didn’t know any bookkeepers and didn’t know much about bookkeeping at all.
A few months later, LinkedIn was kind enough to drop a one-month trial of LinkedIn Premium in my inbox. I signed up at the start of February, selecting the top-tier package (“executive”, I believe), which allowed for 25 inMails each month.
I then went about the work of finding some bookkeepers to solicit for problem interviews (to be conducted over the phone). I started by sending inMails to 15 bookkeepers, selecting them based on their level of engagement on LinkedIn (relatively complete LinkedIn profiles, participation in bookkeeping groups on LinkedIn, substantial connection counts).
The next day, I received my first response – in which I was kindly told that my timing wasn’t good, as most bookkeepers were neck-deep doing tax return preparations. D’oh! (Of course, she was correct. This was early-March.)
I decided to move ahead, anyway. My mental outlook toward the project at that point was what I would characterize as “skeptically hopeful”. I decided that, if the problem/pain was legitimate enough, I’d receive interested responses regardless of the bad timing. Therefore, the next day, I sent out 7 or so more inMails to prospects (bookkeepers who, for various reasons, I suspected would be a little less likely to respond than those in my first batch of 15).
That still gave me a few more inMails to use before the month was up. Somewhat surprisingly (and frustratingly), at least to me, I seemed to run out of promising prospects quickly. With my final batch of inMails, I was contacting bookkeepers without profile photos, with sparse background info, and, in some instances, who chose to hide their name. (Why would a small business owner hide their name on LinkedIn??)
Anyway, I was pleased with my inMail response rate. If I remember correctly, I got 11 or 12 interested responses. This despite my bad timing.
Here’s the inMail I was sending out:
Bookkeeper expertise needed by bookkeeping-naive software developer
[First name of recipient],
Based on your extensive background in bookkeeping (given what I see in your LI profile and on your site – [bookkeeper’s website]), I think you might be able to provide some insight on a product I have in development.
The product (code-named “Project Slip”) is targeted for use by bookkeepers to detect fraudulent credit/debit card transactions for their clients/companies.
My request is for you (and/or whoever you think is appropriate) to review the product and help me determine if it has merit – and to what degree such a product would prove beneficial to your bookkeeping business/career. If I’ve piqued your interest and you’re interested in helping, please reply back to me with your email address and I’ll provide you more detail on the product itself. Eventually, ideally, I’d like to get you on the phone to hear your thoughts (no more than 15 minutes).
Thank you in advance for your time and expertise. If you find this message poorly-targeted, I apologize.
To each response, I followed up with a longer email that provided more detail about the software, what I was hoping to get out of speaking with them, and suggested dates/times for a phone call. The response rate for this follow-up email wasn’t very good. This was either because of, again, the poor timing of my request or because none of the recipients identified with the problem/pain I was targeting.
Regardless, I was able to get two very nice and forthcoming bookkeepers on the phone. From these initial conversations alone, I realized that the initial problem I was targeting was probably doomed to failure. Fraudulent/erroneous transaction detection was a very ancillary nice-to-have, not a significant pain. However, there appeared to be a need for software that would simply speed up the process of receipt-to-transaction reconciliation. I promptly switched my focus to this as my target problem.
However, I had no one else lined-up to interview. I felt a need to wait until after the tax return filing deadline (April 15th in the U.S. and April 30th in Canada) to move forward with my conversations. So, in the meantime, I turned to my social networks for leads. I posted the following on LinkedIn (200+ connections) and Twitter (190+ followers):
Anybody want to refer a bookkeeper my way? Just want to chat. Looking for expertise & to gauge demand for a possible product.
I also posted an expansion on this to the Lean Startup Circle discussion group.
Total number of bookkeeper leads forwarded my way from those posts: two. And one of those two was a self-forward. Hardly impressive numbers.
Finally, the tax return filing deadline passed in the U.S. and I followed up on those 10 or so un-tapped initial inMail responses. Unfortunately, those leads had grown cold and they went nowhere. Dead end.
So, I turned to a couple bookkeeping-specific groups on LinkedIn. I posted the following question on the most promising group:
What percentage of your (professional) time would you say you spend on reconciling (checking-off) expense receipts against bank/credit card transactions? A lot? A little?
Is this something your clients expect of you? Or are they pretty lax about accounting for every charge to the company?
This post sparked a healthy and very helpful (to me) discussion. I figured that these respondents would be good interview candidates. However, I shied away from going right at them, right away. I was wary of making a bad name for myself in such a highly-visible bookkeeping group. So, before pursuing those leads, I decided to cold-call some local bookkeepers.
I hopped on superpages.com and put together a list of 20 or so bookkeepers to call. I then sat down one weekend and dialed, dialed, dialed. (I did it on the weekend because I found it easier for me, mentally, to know that I’d probably be leaving lots of voicemails and getting called back later – rather than having no clue as to whether someone would pick up.)
Here’s the voicemail message I was leaving:
I’m a software developer over in [location near you] looking for some bookkeeping expertise. Please give me a call back. If I don’t pick up, please leave me a voicemail. Thanks.
While this did yield a few helpful conversations, a significant number of the listings were disconnected or defunct. Additionally, I kept on finding myself speaking with accountants instead of bookkeepers – accountants who had little interest in the problem.
I was becoming frustrated because I was finding it difficult to effectively pre-qualify my prospects. While everyone I spoke with offered “bookkeeping services”, that seemingly meant something different to every person with whom I spoke.
I kept looking and stumbled on BookkeepingHelp.com and their pay-for-placement list of bookkeepers. Even better, each listing included a blurb about services provided and, for most of them, a website. I focused on the ones that mentioned reconciliation work and put together a similarly-sized list of bookkeepers to cold-call. However, with these voicemails, I was aggressively more specific with my topic of conversation.
Here’s the revamped voicemail message I was leaving:
I’m a software developer looking for some bookkeeping expertise. Specifically, I’m looking to speak with a bookkeeper that, on a monthly basis, does the low-level work of matching expense receipts against bank/credit card transaction statements. If this is a service that you provide, please give me a call back. If I don’t pick up, please leave me a voicemail. And, if possible, I’d like to speak with the person that actually works with the receipts. Thanks.
These calls yielded a number of enlightening conversations. One theme I kept hearing was the one of bookkeepers refusing to do the receipt reconciliation work themselves – deciding, instead, to push that task off to their clients. When I asked if they thought that their clients were actually doing this work, they answered with a mix of “I don’t know” and “probably not”.
After many, many conversations, I came to understand that the task my software was designed (built) to address was actually an unnecessary task. In short, the IRS does not require each and every business expense transaction to be accounted for with a receipt. Only “grey area” transactions require a receipt. For every other transaction, a statement is enough.
On the flip side, for many bookkeepers that still insist on matching every transaction, those ones I spoke with didn’t see any significant flaw with their manual process.
I made one last switch in focus to Canadian bookkeepers, speculating (based on my first bookkeeper conversation) that the CRA (Canada’s version of the IRS) was more strict. But, after a few conversations, I realized that the rules were basically the same for both the CRA and IRS.
After speaking with ~25 bookkeepers, I’m guessing I had 2 solid (prospective) customers in the pipeline. (But, admittedly, I never broached to them the topic of cost.) While, as I understand it, this isn’t a bad conversion rate for cold-calls, I was left unenthused, for many reasons:
- The pain my software aimed to address was only being felt by bookkeepers who felt that pain due to an inaccurate understanding of the law. As one bookkeeper said, “that’s old-school thinking”.
- For the “grey area” transactions that DO need receipts, alternative solutions such as Expensify targets/handles them very well.
- Most bookkeepers don’t tend to network with or self-reference each other (bookkeeping LinkedIn group members excluded). They work with clients and accountants, but not other bookkeepers.
- I couldn’t find any earlyvangelists (as defined by Steve Blank). Even among the bookkeepers that did the reconciliation work on which I was focusing, very few saw themselves as having a problem (much less were looking for a solution, had tried to cobble together a home-grown solution, and/or had a budget for a solution).
- My anticipated marketing channels were too manually-intensive (expensive) given what I anticipated to be a fairly modest monthly price point (less than $50 each month). Being that nobody seemed to be actively looking for my solution, I didn’t see how Google ads or LinkedIn ads would be effective. Effectively pre-qualifying leads for cold-calls was proving difficult and word-of-mouth didn’t seem like a viable option for the market.
If anyone is interested, here are three progressions of a business model canvas I created as I worked through my Customer Discovery. They are simply labeled by date:
I’m not ready to toss this idea in the trash, but I do intend to put it up on the shelf, possibly for later consideration. (See Dave Troy’s post about Idea Gardening.) Some possible avenues for exploration:
- The small business owners to whom many bookkeepers push off the task of receipt reconciliation, maybe target them? I haven’t pursued this, as bookkeepers are reluctant to offer up client info. Thinking differently, bookkeepers could use my software and turn the now unwanted receipt reconciliation task into a high-margin service offering. But, nobody saw that business opportunity. Anyway, I find it hard to get past the assertion that the task my software is designed (built) to address is actually an unnecessary task.
- I never observed any bookkeepers working – in their natural habitat. Doing so could uncover different business possibilities. However, the domain (bookkeeping) doesn’t interest me enough to keep digging.
- There might be some countries out there (other than the U.S. and Canada) that do require each and every business expense transaction to be accounted for with a receipt. (If you have any knowledge of receipt reconciliation and non-U.S. and non-Canadian bookkeeping, please let me know in the comments. Thanks!)
Even though this customer discovery expedition has, so far, failed, I’m fine with that. I’ve got something built that I, personally, find useful – and nothing more. It took time to find interviewees and actually conduct the interviews, but I figure that was the minimum amount of work that needed to be done to evaluate the viability of my idea. While, from a customer discovery perspective, it has been a failure. The early onset and inexpensive nature of that failure equates to a customer development success.
Please leave any thoughts, advice, or input you have regarding this endeavor. Are there any mistakes you feel that I made? Anything to learn from for my next idea and discovery phase? I’d love to hear them.