如果不能正常显示,请查看原文 , 或返回

The Journey of Selling Sifter

I recently sold Sifter, a recurring-revenue, software-as-a-service business with monthly subscriptions ranging from $29-$149. For almost eight years, I poured myself into Sifter and built it into a healthy business that comfortably supported me. I paid myself well and worked reasonable hours, but it had come time for me to move on for personal reasons. This is about why and how it happened.

After I had been working on Sifter for about six years, I had a minor ankle surgery that unfortunately led to many more surgeries. While I was pleasantly surprised how well recurring-revenue software supported me through the ups and downs of surgeries and recoveries, the process took a toll on me, and having a second child changed my priorities. Without Sifter, the journey would have been much more trying. However, Sifter unquestionably suffered because of the downtime.

The surgeries themselves weren’t really the problem. The bigger problem was all of the appointments and physical therapy sessions. Between the time spent at the various appointments and the time driving to and from, my days were swiss cheese. Finding a decent block of time to get in the zone was incredibly difficult.

In October of 2013, during my most difficult surgery, I spent a full week in ICU and two more weeks in a regular hospital room. Towards the end of that stay, I reached out to Chris and Natalie Nagele, founders of Wildbit, about some of the ideas circulating in my head. I felt like I was out of gas and couldn’t keep running Sifter. Something had to change.

We had a few conversations and briefly explored the idea of me joining Wildbit, but there was too much uncertainty on my end. I wasn’t quite sure what I would do with Sifter, and I still had a ton of recovery ahead of me. I had no idea what my future was going to look like, and burdening the Wildbit team with the unknowns of my recovery didn’t feel right. We decided it wasn’t the right time.

Fast forward a couple of years and a few surgeries, and I was once again questioning whether I had any Sifter left in me. There were a few differences this time, though. I was more or less past the worst of my recovery, so the uncertainty was diminished. We had also added another another daughter to our family. And Patrick McKenzie had just sold Bingo Card Creator. That was the event that got me thinking that maybe there was a market for an application like Sifter. I spoke with Patrick briefly, and he highly recommended that I talk to FE International (or “FEI” from here on out) about selling Sifter. The idea of any kind of third-party agent involved in the transaction made me incredibly uncomfortable, but I had a lot of faith in Patrick’s recommendation.

Another thing I learned that’s worth noting here is that depending on the size of the business, the terminology for the listing agent varies. For smaller SaaS businesses, you’ll probably hear the term “business broker” or “website broker” simply because the business value is in the single-digit millions or less. Above that point, you’d probably hear them referred to as “investment bankers.” I’m not quite sure why there’s different terminology, but my impression has been that they’re both doing the same type of work on different scales. “Broker” seems to carry some baggage with it, and on some level, it probably created some of initial trepidation for me. I’ll use the term “broker” throughout here, but based on my experience with FEI, “advisor” feels like a better fit.

Despite Patrick’s recommendation and after some superficial research on my part, saying I was hesitant would have been an understatement. Did I really want to move on from Sifter after almost eight years? Would it really be possible to find a buyer who wanted to take care of Sifter rather than exploit its customers? Would that person really be willing and able to pay an amount that would make the process worth it? Would working with a broker be a terrible experience?

With those questions in mind, I reached out to Chris and Natalie again to let them know that I was contemplating moving on from Sifter. I wasn’t planning on making any moves at that point, but I wanted to give them a heads up that it might happen. Within hours, Natalie and I were talking seriously about me joining Wildbit. Of course, I had to figure out what to do with Sifter first, but we agreed that it was worth finding out if I could sell it.

That day, I reached out to FEI and spoke to Thomas Smale. I went into that conversation with every reservation possible, and even with Patrick’s recommendation, I expected a silver-tongued sales pitch telling me whatever I wanted to hear so they could list Sifter. Everything Thomas told me almost sounded too good to be true. Much to his credit, the end results of selling were almost exactly inline with his guidance, but we’ll get to that later. Throughout the sales process, I worked with David Newell at FEI, and it has to be said that he did an incredible job facilitating the process.

So after some soul-searching, discussions with my business partners, Wildbit, and my wife, I decided that moving on from Sifter was the right thing to do. We sorted out my role at Wildbit, how the process of selling Sifter would work with me working full-time at Wildbit, and when I could start. From the moment I started thinking about selling Sifter to the point that I joined Wildbit was maybe a week, and that’s when the real work started.

If you’re familiar with Sifter and Wildbit, you might be wondering why Wildbit wouldn’t just acquire Sifter, and that’s a great question. We did discuss it, but given Sifter’s size and focus, it simply wasn’t the perfect fit that it would need to be. Plus, Wildbit already had enough products to manage. Adding a smaller one like Sifter to the mix would have added more complexity and distraction than value. Sifter just wouldn’t have had as bright of a future within Wildbit.

Alternatively, you might wonder why not just leave Sifter in maintenance mode and cash healthy dividend checks indefinitely. There are several reasons, but first and foremost is appreciation for Sifter’s customers. It has always been important to me to take care of customers, and putting Sifter into maintenance mode didn’t feel fair to them. The other major reason was focus. Even though Sifter doesn’t generate a lot of support requests, it does generate some. And any application needs regular updates if only for security reasons. So keeping Sifter, or any app, running, secure, and healthy requires a non-negligible amount of attention. Long story short, Sifter’s customers deserved better, and I didn’t want to do something halfway.

And so I set out to find Sifter a new home.

When you’re selling a software business, there are quite a few considerations to keep in mind. So much of the business is intangible, it presents challenges for buyers to validate claims made by the seller. As a result, any buyer is going to have a lot of questions, and they’re going to want historical data and proof of everything for the last 12-18 months. Pulling all of this information together is involved, and this is the part where good brokers start to earn their keep.

The premise of a good broker is that while they take a percentage of the sale price, usually 10-15%, their expertise helps you get a higher total valuation for the business by saving time, presenting the business in its best light, and reaching more buyers. After working with FEI to sell Sifter, I highly recommend working with them. I’ll get into the details here shortly, but for now, suffice it to say that they know what they’re doing, and they’re good people.

So what does selling a company look like? I can only go into what I learned with Sifter, but Patrick McKenzie has written extensively about the process of selling his application, Bingo Card Creator. His experience and advice closely mirrors that of my own, but it’s definitely worth reading for his perspective on the process.

With Sifter, the entire process from starting with FEI until we closed took a little over six months. However, that’s only because we had to walk away from the first deal on the closing day, and we had to begin again looking for a new buyer. Had that deal gone through, it would have been about four months, but I’ll get into that later.

The Listing & Prospectus

The first phase of selling is preparing the listing. Any buyer will want details on revenue, site traffic, search engine traffic, news stories, your road map, the team members, operating costs, personal information, and much more. Since software isn’t tangible like a traditional business, it needs a lot of documentation to illustrate that it’s not all smoke and mirrors.

Unless you’re extremely experienced, you won’t think of half of the things that buyers would want to see before they even express interest. You might even be tempted to hold back on some information that you feel isn’t flattering, but that would backfire. If it doesn’t come up right away, it will eventually come up. Better to have a potential buyer be turned off by that detail before spending too much time with them and then walking.

From my research, this is one of the areas where FEI excels. The questionnaire they give to the sellers combined with the list of documents they ask for up front is incredibly thorough. While it’s tedious to gather all of this information up front, it’s something that you’ll have to do at some point regardless. By gathering it up front, you can help narrow interested buyers down to really interested buyers much more effectively, and that ensures the later parts of the process run much more efficiently.

After FEI gathered all of the information, they compiled it into a 21-page prospectus covering company history, our reasons for selling, market trends, opportunities, traffic, financials, operations, maintenance, support, all of the related domains we own, and more. These all focused on the last 12-18 months and included raw numbers, graphs, and screenshots of graphs. Most buyers will want to see at least the last year of detailed financial history to get a feel for the direction the business is trending.

One of the key tasks that FEI had me complete was an in-depth questionnaire about the business, related assets, day-to-day operations, support, customer demographics, competitor analysis, marketing history, email lists, SaaS metrics, billing, and so much more. In total it was well over 100 questions and took about four or five hours to complete. Completing the questionnaire was the point where I really started to recognize just how little I knew about the information buyers would want.

At the end of this process, you’ll have two key numbers. The first is SDE, or Seller Discretionary Earnings, and the second is a multiple. These two numbers work together to give you an idea of what your business is worth.

SDE represents the amount of money that’s left over after the business pays its mandatory expenses. The biggest lesson for me was that your salary as the owner is generally considered discretionary if the business can operate day-to-day without you. So, if you have a business making $100,000 per year in revenue with $15,000 in expenses, $80,000 in salary and health care, and $5,000 left over, your SDE could be as high as $85,000. A new buyer might have to hire someone at a lower salary to handle some of your responsibilities, and that would likely decrease your SDE. So, let’s say the final SDE ends up being $70,000.

The next key number is your multiple. With SaaS and recurring revenue, valuations will generally be between two and three times SDE. While there are quite a few factors that influence the multiple, determining it is more art than science. Is the business shrinking, growing, or stagnating? How steady and reliable has it been over the last 12 to 18 months? How old is it? What’s the churn and average lifetime value of customers? How complicated is the business technically? How good is the codebase? Is there any significant technical debt?

So, with SDE of $70,000, your business would reasonably be worth somewhere between $140,000 and $210,000. A good broker would give you a narrower band of expectation and be able to give you a more precise multiple. Of course, at this point, it’s just an estimate, but with a good broker, it should be a fairly solid estimate.

The other factor that’s likely to affect the multiple is the deal structure. An all-cash deal would mean a lower multiple, but a deal that’s a mix of cash and financing would generally get a higher multiple since the payments are being drawn out, and the new buyer is getting to partially pay for the business out of its profits. I’ll talk more about deal structures later.

Finding a Buyer

Once the prospectus is complete and you have a target selling price, the next step is listing the application and searching for a buyer. In the case of FEI, they do this by circulating the prospectus to their targeted list of potential buyers and gradually expanding the circle to a wider audience of buyers if needed.

The ability of a good broker to discreetly reach a significant audience of potential buyers is priceless. Not only do they reach a wide audience, but they help screen potential buyers so you don’t end up wasting time with unqualified buyers.

Based on conversations with FEI as well as other friends who have sold their businesses, publicly marketing and selling a business leads to an incredible amount of tire-kicking. Some people may have no intentions of buying. Some may simply not have the resources. And yet others may just be fishing for data for competitive reasons. In all of these scenarios, you end up wasting a lot of time with individuals who simply aren’t serious.

Beyond saving you from endless conversations with unqualified buyers, a broker can also help screen serious buyers for which your business may not be a great fit. All of this saves you time and ensures that by the time you’re talking to a potential buyer, they’ve been vetted and won’t be wasting your time.

With Sifter, I took calls with five different people, and we received four verbal offers from those five, and two of those went on to a letter of intent. Just before we signed our first letter of intent, we also received another verbal offer from someone with whom I hadn’t had a phone call yet, but we decided not to pursue that because we were comfortable with the letter of intent that was about to come across. I feel like those numbers speak for themselves. FEI did an excellent job presenting solid offers from serious buyers.

Letter of Intent/Interest

A letter of intent of is a simple document that states the agreed upon price, closing period, and terms of the deal before due diligence begins. Unless something materially adverse comes up in due diligence, the final deal structure will match the terms in the letter of intent.

That deal structure can vary widely and come in many different forms using a variety of considerations. First, and most common is cash, and as they say, cash is king. The more cash involved in the deal, the lower the price will be. In some cases, the deal will involved a holdback. This is just some portion of the cash that sits in escrow for period of time, usually a year, to help cover any unforeseen or unreasonable costs that arise due to any failures on the seller’s part to disclose problems with the business.

The next most common facet of deal structures is seller financing, where the buyer pays for a portion of the business over the years after purchase. The downside of this is that as a buyer, you’re not going to see the money for a while, and the buyer is paying for the business out of its profits.

Another common method is for the buyer to pay for a consulting agreement. While it’s common to include some amount of hours over the first three months for transition assistance, buyers can also set up consulting agreements to keep original team members on retainer for a longer period of time. The fourth common element is to tie some portion of the deal to the non-compete provision.

One other key part of a letter of intent is that it will also stipulate exclusivity for the buyer. This helps protect the buyer to ensure that as a seller, you aren’t continuing to shop the business and look for better offers during due diligence. Since due diligence requires a significant investment by the buyer, it would be unreasonable if another buyer could come in at the last second and derail the deal. Think of it as a sort of engagement period for businesses.

Since we walked away from the first deal, we ended up with two letters of intent. While that wasn’t ideal, it was a great experience to compare and contrast the approach of different buyers. I won’t get into the details of the first deal falling through because the reason wasn’t really anything juicy. An issue was discovered the day before closing, and we couldn’t reach an agreement with that first buyer about how to handle it.

This was the definitive moment in my mind when FEI went from a broker that I respected, to one that I completely trusted. With the first deal, FEI was literally a couple of signatures away from being paid for all of their work. They had every reason in the world to encourage us to concede to the buyer’s request for this issue. Our choice to walk away from the deal would mean significantly more work for them to find a new buyer, a lot of work with that buyer, and a good chance their paycheck would be smaller as a result. However, FEI fully supported my decision not to concede. I never once felt pressured in the slightest to accept that first deal.

For context, the issue at closing wasn’t the first problem with the deal. There were multiple concerns throughout the process. The buyer’s first offer was an incredibly low ball offer and only increased when we received another offer. Then the letter of intent was materially different than the verbal offer, and they knowingly made the change without talking to us about it ahead of time. So that deal started off on the wrong foot, and at each step of the way, we conceded everything they asked. So the issue on closing day was simply the last straw.

With the subsequent, and ultimate buyer, we were incredibly up front about the issue, and, while it was a minor concern, it wasn’t something that he felt was important enough to merit not buying the business.

This is another spot where I can’t speak highly enough of FEI’s experience and counsel as we navigated the process. They helped me understand which behaviors and requests were perfectly normal and which requests were a bit unusual. That experience and guidance played a large part in my education and feeling comfortable standing our ground on the final issue that killed the first deal. More on that later.

Due Diligence

Due diligence is the phase of a business acquisition where the buyer goes over everything with a fine-tooth comb to make sure that all of the claims in the prospectus are true and that there aren’t any other problems that will come back to haunt the buyer.

With a software business, you’ll effectively have three categories of due diligence: operational, financial, and technical. Any informed buyer is going to want to see real financials, proof that the numbers are real, and proof that proof is real. They’ll also want to understand operations. Much of that is related to financial, but this will also include things like the processes for support, customer refunds, and other regular business tasks.

In our case, thanks to FEI, the vast majority of the financial and operational due diligence was handled when they originally created the prospectus. So the only financial due diligence left on our end at this point was primarily about getting all of the PDFs and receipts that proved the numbers were accurate. We had a few calls to answer questions, but they all went very smoothly due to the advanced preparation FEI guided us through.

To me, the technical due diligence was much more interesting. I don’t remember the exact amount of time, but it involved hours of screen sharing to dive into the code and tools and processes for everything. What’s the deploy process like? How is everything monitored? How’s the test coverage? What’s the overall code quality? What other applications does it rely on? Where is it hosted? What has the uptime and response time traditionally looked like? Are there obvious opportunities for performance improvements?

With Sifter, code quality was always something I was concerned about. When I launched it, I cut a few corners, but over the years, I constantly chipped away at our technical debt. It wasn’t perfect when we sold it, but we’ve heard nothing but positive things from the buyer. With smaller apps and teams, it’s more common for things to be less organized since smaller teams can get away with it a little more easily. You should always be investing in improving your code quality and documentation, but if you’re serious about selling, it’s even more true.

Asset Purchase Agreement

At the end of due diligence, assuming the buyer still wants to move forward, you’ll begin hammering out the details of the APA or Asset Purchase Agreement. Barring any unusual discoveries by the buyer, this will take the deal structure from the letter of intent and add details like a non-compete, intellectual property, and other key details.

This was another part of the process where our experience with the first buyer was night and day different from the final buyer. The first buyer had his lawyers draft the APA, and it was long. Our lawyer’s counsel was that it felt far more complicated than a deal of our size merited. It required a fair amount of negotiation on the finer points, and the exhaustion from going through that process and conceding most points contributed directly to my willingness to not give in on the point that killed the deal on the closing day.

With the second buyer, we had been introduced by a mutual friend, and we had several more mutual friends. The entire process was underscored by a much higher level of trust and respect. As a result, things moved much faster than with the first deal, and there was very little back-and-forth. More importantly, it was an all-cash deal that prevented the need for negotiating a bunch of details in the deal structure.

Closing & Transfer

Most closing processes will involve escrow of the cash and domain names until both sides have fulfilled their obligations. With a web application, there’s a whole lot of account and domain transferring going on. Email accounts. Domains and DNS. Source control. Hosting. Exception tracking. Help desk. Monitoring. And a dozen more. We spent a couple of multi-hour screen-sharing sessions to hand these over, and things went pretty smoothly.

As I mentioned earlier, it’s generally expected for the seller to remain available via email, and occasionally phone, for the first few months of the transition. While I’m sure some sellers are dying to move on, I may have been a little more involved than the new team needed me to be because I wanted to make sure they’re happy and that everything was incredibly smooth and transparent for customers.

For the most part, I’ve only been adding notes to support requests as they come in. These vary from adding context and insights to feature requests to just helping share where to find the answers to certain questions. Occasionally I’ll help troubleshoot with a few of the more obscure areas of the application, but that has been pretty rare.

Life after Sifter

I knew it was the right time to move on from Sifter, but I did wonder throughout the process if I would miss it. The grass is always greener, right? I’ve never poured so much of my life into a single project, and bug and issue tracking has been something near and dear to my heart for long before I started Sifter. There are certainly some aspects that I miss, but overall, joining Wildbit has been a fantastic decision.

Recently the new Sifter team, JD and Colin, happened to be spending some time around the corner from me. We found some time to meet up on a nearby patio for drinks, and it felt like hanging out with old friends. I don’t imagine most acquisitions end up feeling like that, but it’s reassuring to know that the new caretakers believe in the same things as me. I could not be happier or more confident about how things turned out.

Advice for selling your SaaS business…

If I had to do it over again, I absolutely would. If you’re in the same spot and considering selling your own app or web site, I have some thoughts based on what I learned throughout this journey.

  1. Talk to a broker early. (Again, I highly recommend FEI.) You don’t have to work with a broker, but it’s not going to hurt to talk to them. It’s more likely that they’ll help you realistically calibrate your expectations.
  2. Listen to your gut. Our first deal fell apart on closing day, but in hindsight, the whole process had a much more legalistic approach that had been slowly contributing to doubts in my mind.
  3. Don’t be scared to walk away from the wrong deal. Again, in hindsight, the closer we got to the end of the first deal, the more eager I was to get it over with. It wasn’t until the impasse on closing day that I finally felt 100% alright about the deal not happening. Of course, there’s no guarantee that another deal will come along, but if you’re unhappy on closing day, you’ll end up with regrets.
  4. Get a jumpstart on getting your books setup so everything is paid on a monthly basis. Pre-paying for services on a yearly basis makes life easier, but it also leads to a few more headaches when buyers are looking at your books. Peaks and valleys from one month to the next creates unnecessary scrutiny. If everything is paid on a monthly basis, it will smooth out your expenses and make it simple for a buyer to evaluate.
  5. Cut unnecessary costs. Sifter was comfortably profitable, and while I was mindful of spending, there were certainly some luxuries that we spent money on. These all made the business look less profitable and generated a slew of questions that could have been entirely avoided.
  6. Invest in documenting and cleaning up your code. You’re going to have to do it for the transfer anyways. By doing it up front, it’s going to put the business in a much better light, and it’s going to make your life easier post-transfer. Even if you’re a solo developer, documentation is wonderful. In hindsight, all of the documentation I did ahead of the sale would have served me well years ago.
  7. Invest in a good clean codebase. Every team cuts a corner here and there. Some more than others. Regardless of your situation, the more you invest a good clean codebase with great test coverage, the easier everything will be. The sale will go more smoothly. You’re likely to get a better valuation, and you’ll spend fewer hours after the transfer.
  8. Get concrete metrics on the quality of your codebase. Things like test coverage, Code Climate, or other static analysis tools can really help when it comes to provided tangible measurements of how good your code is. They won’t be the only metric buyers will look at, but they help.
  9. A reliable deploy and server provisioning process falls under the category of a good clean codebase, but it’s important enough to call out separately since release management is frequently one of the messier “processes” a team can have. Make sure that it’s simple, resilient, and fully automated.