On Starting a Software Business
stephaniehurlburt.com
stephaniehurlburt.com
One pragmatic strategy to find product/customers:
1) Pick an industry
2) Ask someone in that industry what they use spreadsheets for
3) Build something better
So I don't think it's a foolproof scheme, unless your solution is also programmable/customizable, which is a lot of work.
Here is one random example that comes to mind. "Risk Management" is something every company needs to do or pretend to do.
So someone gets stuck with this responsibility and they create a spreadsheet that looks like: Department | Risk | Likelihood | Impact
So this is a fairly mediocre solution but it kinda works and "risk management" for "Small Company Co" becomes "Bob's thing".
Of course there is very likely software that already does this but in my experience it's usually what the hipster programmers call "enterprise garbage". You know what i'm talking about - buzzwordy site with stock photos, long annoying sales cycles, no online demo or sign up, some java server you have to deploy, and some extjs/jquery interface that is ugly and each click takes 5-10 seconds. You aren't competing with these guys, you're going after the long tail of small companies they are ignoring.
Your move is to register some fun name - "Riskly", "Riskque", "Risksalot", etc. Next, hire a graphic designer to put together a fun colorful site, hire a college kid to make a fun video with royalty free upbeat music and narration showcasing your app, and build a simple CRUD app with maybe some built in risks for certain industries and some cool simulations/reports.
So go spend a few years of your life and see if you can't 'Disrupt/Fix Risk Management'. Or go work for a large company for eight years, take all your earnings, go to Vegas, and let it all ride on a 10:1 bet. It works out about the same.
HR, benefits, invoicing, inventory tracking, project tracking, etc etc
And the fact that there are business providing most of those testify to how good of an idea it is (and how crappy the existing MBA-Sales enterprise solutions are).
While we realized it before starting, we didn't fully grasp how much you need people with "computer literacy" to build these things, for lack of a better term. Just thinking about the schema of the application you need is something a lot of people can't really do yet.
As far as I can tell it's got very little to do software intuitiveness, as I've noticed the same effect on phone calls with prospective customers.
Even as enterprise moves towards managed SaaS platforms, there'll always be a huge place for power-users. Much of the time, even just defining the problem that needs to be solved takes one.
The CRUD application is definitely not the hard part, but it can be the long part.
Or reach out to ethbro . coAtGma il.
Yep, almost everyone you casually talk to about a problem domain is only casually talking about it. They'll dream, but won't let you know about the pitfalls.
After you build something, the next questions are usually, "Can you change this just for us?", which is a fun task if you have 9 other customers that won't want that change.
I've noticed this pattern in both business and in individual engineers.
You either have a spectacular product that theoretically could sell itself with initial exposure, or you have the best marketing and a product that might as well be vapor.
Usually it's the mediocre product that's somewhere in between that wins.
In engineers I've noticed that some of the brightest I've worked with are unsociable, or have no self esteem, and get stuck in the same place for 20 years, and on the other side of the spectrum are talentless hacks that are great at selling themselves and moving around enough to increase their salary.
Meanwhile I feel like a fraud because I'm not the best, but good enough to recognize my worth and sociable enough to put myself out there.
I leverage the shit out of that and feel somewhat guilty about hoping that the really talented local engineers don't catch on to the idea.
Younger company with unknown decent product? More heavily spending on sales.
Older company with industry recommendations from previous clients? Spend less on fewer, more technical sales people.
Older company with no recommendations from previous clients? You're doing something wrong, and sales isn't the place you should be spending your money.
I feel like normally this logic is inverted at the C-level though. Sales are low and product isn't getting traction? Hire more sales and marketing folks and put off those engineering hires.
Consulting requires linear increase of resources. Software product requires logarithmic increase of resources. Therefore product business is more efficient to scale than consulting.
Accenture is #493 of the S&P500, at $101k of revenue per employee. (!) They're doing a bit better than McDonalds (#497). The scalability argument holds -- compare to Netflix, which scales quite well (#23, $1.9M/employee).
When people say "it scales" they are largely referring to the cost of serving incremental customers. Software does that well, consulting doesn't.
And then ideally via the consulting you find a niche(s), build your own processes & tooling for your niche to deliver value to clients, and then can product-ize once you actually understand the space and the market, who the customers will be and how to reach them, because you're already doing it via the consulting and just need to scale up.
It also serves as a sanity check that you're actually building something that has value, because you prove it to yourself via invoices for your work rather than investment in an idea.
Talk to the users, try to find out how many others are also using it. Rewrite it in modern technology, sell it back to all the current users (and ideally, also sell to others in the same line of business who might need it).
This is exactly what I did!
On top of that consulting opens door, you get very valuable insight inside certain industries and get paid for it.
Also, I'm sure I'm not the only one who is not interested in the slightest in building a unicorn or dealing with investors/VCs.
Building software with no clients requires capital, I'd rather keep mine for my family rather than burn it in the wind into something that is most likely to fail.
Do work for customers. While doing work, try and find a shared pain point across your industry/customers. Build software for it and use your existing relationship to sell it. Or sell it and then build. Whichever.
Your advice is spot-on, although I would likely CAPITALIZE the words "build something better". Not so easy a proposition.
I started and 10 years later made a successfully exit from my company, we never raised any money, the above figure is what it took us to go from financing our development our self (by having two jobs), to having enough customers to break even.
I think everyone should, bootstrap one company to get a feel for how it is, will make you a better entrepreneur.
I launched Atlas in 2007, just one year prior to the economic crisis that followed. Nothing ensures that you keep an eye on every bill that is paid and focus ruthlessly on efficiency gains more than a global financial catastrophe. Getting through that taught me I was capable of running an already tough business under pretty much any circumstances, and gave me an incredible confidence boost.
Not that I'd recommend it, definitely some of the toughest years of my working career.
Currently, this is a manual process, paper driven with digital photos, and after wards, somebody back in the office would transcribe the hand written notes into an excel spreadsheet along with pics.
This process could be made easier with an automated app on the phone. I had thought about making such an app, but haven't yet.
I guess when you talk to someone, you naturally gravitate towards issues they have.
Where is your contractor friend based? I am looking for a contractor, or some contacts, for a job now.
That really depends on context. Very often something starts out sounding like "document with text and pictures" or "database with some tables in it and some forms for UI", making it an easy job with generic business software. However, often when you look at the details the actual requirements are more complicated and require some level of domain-specific functionality.
There are many applications that are aimed at a certain industry and provide some core function that seems quite generic but then some other more industry-specific functions on top. That certainly includes tracking, displaying and annotating specifications and inspection results in the construction industry, where there are multiple players in the market already.
disclaimer: I met the founder once at a party.
We work with a lot of clients in manufacturing and construction. Our system has a way to build and schedule regular inspections, notifications, etc. We're working on business process management and moving into regulated industries.
A friend of mine works in sewers for a big construction contractor. I want this company using our stuff so that his job is more safe. Right now they take readings from his sniffer and write them out onto a sheet in a log book and throw that data in the back of the van. It might be three months or so before someone enters that data into a spreadsheet. If there was an incident and my friend happens to not make it out of the hole that day -- he's just another statistic. I hope that by taking a reading with our software we can at least anticipate the environment he's diving into. His co-workers won't be able to scribble some non-sense on paper and throw it in the back of the van before he dives in -- they'll be accountable and people will be notified if they don't.
It's amazing how much our customers still use paper and spreadsheets despite having SAP, Oracle, SAS etc servicing them with solutions for that last couple of decades. Turns out having a small team that can move fast and listen to users on the ground floor is still a good strategy.
That's step 1. Step 2 is give them some solid advice in the form of a few things to try. Say "Well here's what I would do" and tell them what you'd do. If they keep asking questions, tell them something along the lines of "I did something like that for another guy and charged him $X/hour, you want me to stop by?"
Mentioning an hourly rate lets them know that even though this might be a hobby for them, its not a hobby for you. It also weeds out the tire kickers.
When people say Bay Area I think they really mean Palo Alto!
https://news.ycombinator.com/item?id=3298905
That poll indicates that the significant majority of HN readers reside outside the Bay Area of San Francisco in the US.
Four years later of 15-20 hours per week side work and that chance encounter had an amazing impact on my life.
They also need to ENVISION their goal, and then WORK HARD, and OVERCOME ADVERSITY.
Also TEAMWORK AND GRATITUDE!
> "give first, take later"
This was when they were looking for leads.
> "don't do work unless you're getting paid for it."
That was in the context of the product they were selling. A choice not to work on the product when there is not a customer paying you to do so. I can imagine they chose to operate like this because it is very easy to sink many hours into perfecting a product that does not have paying customers. I believe they will step away from this rule later when they have so many customers that they want to develop ahead of payments.
I would also cut off the advice giving at a certain point and bring up money.
I helped a local business group with their website, their previous host shut down and all they had was an archive on CD, they had some quotes for 10-20 hours to get it back up and running (and then everyone wanted to charge them hosting/domain registration).
I did the work for free. A couple hours importing databases and site files, configuring WP/Joomla (I forgot which), some testing... My hope was that I'd get a few leads from the 50-60 companies in the business group.
I was contacted by a few but once they found out I wouldn't do the work for free...
In my experience, there are really only two levels of pricing that work. You can work literally for free, or for some token payment so someone is an official client of your business. This is sometimes helpful very early in a business to start some sort of portfolio and establish a bit of credibility, but otherwise I think it should be reserved for genuine charity work where it's clear that it's a gift and not your normal working practice. Everything else should be charged at decent professional rates immediately, both so you don't undersell yourself and to dissuade potential clients who can't or won't pay decent professional rates and all the other similar potential clients in their network.
Agree. Keeping cheap and non paying clients AWAY should be a top priority for any business. In consulting, that's achieved by ALWAYS charging and being clear about your rate.
Which is to say, if I can solve your problem for you before I get tired of standing up, I'm not going to bother to charge you. In my case that's usually a half hour or so.
People seem to respond well to that, in general. I think it's in line with how most people view the value of their own time.
It's also a one off, as well. Just because it only takes me 15 minutes a week doesn't mean it's free.
[1] Like charging a startup $50 instead of $150 an hour "just until they get their funding".
You probably can't. If you perform some regular work for someone for free, then they will think that that is your rate, and they come to you because your rate is so low.
The help the OP is talking about, is things like:
"Our most successful publicity attempts always sprung out of helping others-- we did several community workshops, mentored lots of people, helped people find jobs, etc. As time went on, we also got more invitations to speak about the product at events and conferences."
I don't attribute this to impostor syndrome.
I feel like if I'd learned web development five years ago I wouldn't know anything about it today.
This is a wrong assumption. Web development just gives the feeling of moving fast, it's not really the case. The foundation of web development is roughly the same as 5y ago. Most things from the http protocol, web servers, data storage, authentication, websockets etc. still works the same. Give or take a few UX and state-management tools and we're not that far into the future.
Also due to the huge investments in the web during the past decade(s) things should get slower because of legacy issues. Think fe. how long it's taking browser vendors to dump Flash or to adopt WebRTC.
You can still be a generalist as long as you put those skills to the service of a specific community. This could be region, vertical, domain, etc. The important part is to be clear who are you aiming at, so you can hustle and leverage the network/reference effect in there.
Many developers believe that conversations "just happen" for some elite breed of networked/famous/etc developer that excludes them. Sales conversations generally don't "just happen", for anyone. You write some pitches and make some calls.
Test whether they'd buy it with google adwords or facebook ads and conversion rate to filling out the form and putting in credit card. Refund everyone.
Imagine how it would be viral and test your hypothesis as well. So your customer acquisitin cost goes down exponentially!
Most people have neither or just the first one.
Writing about "Starting a Software Business" which boils down to "just know the right people" is a bit demotivating :\
I've also read a bunch of articles that suggest unequal splits, and the reasons they cite seem to boil down to the idea that the split should reflect the proportional risk taken, and not much else. There are times when founders truly take the same amount of risk, and an even split is justified under the common VC thinking.
In my limited experience pitching to VCs, as someone in an even-split startup, I believe VCs do prefer to have a primary person responsible who is in the CEO seat and has authority to make unilateral decisions. I can imagine an uneven split makes it clear when that's the case. I'm certain many of them have seen cases where even split partnerships were made more difficult and ended up being unfair largely because of the even split. I'm also sure that there are cases where it's harder for them to communicate and negotiate with multiple founders rather than one.
Those reasons don't matter if you're not taking funding, so even split in a contracting partnership (even split effort), turned profitable product startup that doesn't deal with VCs makes a lot of sense.
Sometimes it's just a conversation founders want to avoid having because it's uncomfortable, but that in itself is a bit of a red flag.
I've also watched and been part of lots of partnerships where the expectation before hand was equal contribution, and the reality after the fact was lopsided. I feel fairly lucky that my most recent experience truly was equal risk and equal effort; my even split experience has been truly reflective of what's happened, and truly fair for both of us.
The big mistake I have made is giving people besides my co-founder too much equity. Nobody else who's contributed to my company so far other than my co-founder have lived up to what they said they'd do, and we gave out too much equity in advance of them doing it.
Considering that, and considering both the issue you brought up and the reasons YCombinator gives for pushing toward equality rather than away from it, I think the main thing founders need to do is have a mechanism for not committing equity in advance of the effort. It's important for founders to be on a vesting schedule so that when one stops contributing or contributes less for a long period of time, you can shut it off. And it's important to communicate that things are going off the rails, and then actually take the action to shut it off.
That's really basic and somewhat stating the obvious, but if you are diligent about cutting off people who don't contribute, then there's no reason to shy away from equal split plans in advance. My experience has been that people I've partnered with for equity aren't avoiding contribution maliciously, they simply and honestly didn't realize in advance that they don't have the time & energy they thought they did.
Even later on, once they have a product, why would you change that? 50-50 is great, since you need both parties to agree to do something, which is essential in a small business like this.
Do you think that the VC money is what drove them to behave in a way that you don't agree with? If so, how do you square that with the fact that they are now publicly traded and not subject to VC whims?
They agree my friend deserves more equity, but can't agree between a 70-30 or 80-20 split, so other founder hires lawyer and holds my friend hostage for a lucrative buyout. Company is now worth significant amount of money almost solely to my friends hard work/investment, but legally he can't move forward until co-founder signs off on every decision.
But outside my example, starting any new idea as a 50-50 partnership without vesting requirements is a terrible idea.
I'm not arguing against vesting, but I don't understand your example.
1) I get that to get the first users, you have to go out and recruit them. That's the advertisement part. So yeah, I get how you get users by advertising.
2) I get that when you have a million customers, they become the advertisement and they are the ones who are bringing you new customers. Your product "ad surface" is way bigger than simple banners through them.
Now what I have a hard time to understand is how 300 users procures the same effect as 2) (or even 3000). I get that they gonna get you 30 more customers by talking about your product to their "friends" who, if everything goes fine, at their turn gonna bring you 3 more customers who gonna bring you 1 more customers, for a grand total of 34 more customers. And that's the end of it. That wont grow much bigger. So basically, you would have to go back to 1) to get new users.
What am I missing? Does a user have to bring you n more new customers, n > 1? How many customers do you need via advertising so that the compound growth effect kicks in?
edit: I can see actually 2 types of compound growth effects: a) a user explicitly recruit another user (ex: "hey my friend, come play to this game"), b) a non user "see" other people using the product and he tells himself that might be a good idea to try it. Anyway, so basically, what I see is that, early on, you can rely only on a) so, in average, your customers have to get you n more new customers, n > 1, by explicitly recruiting them for you. Which, if Im correct here, would give an idea how badass the product must be.
The long answer...
New customers come from two primary sources: 1) marketing and 2) referrals. Marketing is things like events (trade shows, media hits, advertising, reviews, etc). None of those things happen without outreach on your part. Referrals are when (as you described) existing customers tell other people and a few of them try your product. This will not be constant, right after a customer buys your product/service, they may be excited and tell some friends. Best to model referrals over time with cohort analysis.
At the same time, you will lose existing customers. Like referrals, this is best modeled as cohort analysis. After a customer signs up, after 1 month, x% will leave, next month, a different %, next month yet another %. Typically the churn rate declines for a given customer cohort over time (though there may be bumps along the way).
So, you have customer inflow and outflow. It would be nice if there was a simple formula, but the marketing efficiency, the referral rates and churn rates will all be constantly changing. So, there is no way to know in advance if or when compound growth kicks in until you have solid data. It will vary by market, company and product.
Ultimately, I was focusing on the fact that if a customer brings you n new customers, n < 1 (and real number), it means you get new customers only via marketing. If n > 1, you get new customers with something else than marketing - so to speak.
Edit - though under 1 you can see a limit. For example, if n = 0.5 then you would expect to grow to twice the size. In reality it's much more complicated, but on a purely mathematical basis this should be right.
So it's the divisor on your marketing expenditure for any given customer quantity.
N > 1 never reaches a steady state, those are "viral" businesses.
K > 1 is a runaway train, hold on. It happens rarely and apparently when it does, it's the ride of one's life. (Apparently, anyhow - I've never experienced it.)
K < 1 is still extremely useful - think of it as a marketing amplifier. So if you're paying $10 per click on Google Adwords, an your conversion rate is 2% for those clicks, you're paying $500 for one customer.
Implement some application flows that incentivize referrals, get that k to, say, 0.25, and now that $500 is getting you 1.25 customers - and your cost per customer has dropped to $400. That allows you to outbid your competitors for those adwords. (Except of course that they're also focussing on virality.)
Im not entirely sure this is that rare. Lets think of it like this: whatever gives you new customers is part of what I call currently your "ad surface" or "marketing surface". If a customer brings you another customer, it's just because that guy is marketing too - obviously. So everything that gives you customers is marketing. But of course, thousands and thousands of customers are so big of a surface that they dwarf any given techcrunch articles (most likely). My point is coming.
Now let's say one of your customer write about your product in his blog. Maybe that gonna brings you, say, 25 customers. Another make a youtube video that makes 500K views. Better than that, techcrunch hears about your product and write about it spontaneously. As we can see, those are indirect marketing surfaces and they can get very big.
Now, let's talk video games, about some niche type of games, for example, roguelikes. Roguelikes arent played by many people and yet some of them are well known from roguelikes players. Ex: DoomRL. And this is my point, it seems not plausible to me that DoomRL got his users and his niche-fame from N < 1. It didnt get niche-famous by direct customer-to-customer referrals either because if you play rogueslikes, it's unlikely your friends does too. Yet DoomRL is famous. So DoomRL had N >= 1 at some point in time and not by direct customer-to-customer referrals.
The way I'm seeing it now is that, most of the time you are in N < 1 land, but occasionally you hit some special customers that completely change the average toward N >= 1. And it happens more than one think - it's counter-intuitive in some ways. Also, at some point, your product is famous, and then customer-to-customer referrals (of type b. see edit of original post) can get big.
Which means, strategically, you need to get day-to-day-N as big as possible obviously, and you need faith: you gonna hit those special customers occasionally that gonna critically change N upward. That's my current understanding. It's getting better I think :)
What is this referring to exactly? Administration fees for starting a company?
I like the people I'm working with, am paid well, have a good work environment and have interesting work to do, but for some reason I'm not happy and feel on the verge of burnout (if it hadn't happened already). I've done startups in the not too distant past and while they were a lot more stressful (and paid very little - we had some funding and had revenues but we never broke even), I was definitely happier working in my own thing. My problem is similar to yours. My personal network is good for changing jobs, but I have no idea how to make it work for consulting work or a startup and I'm really terrible at networking and find it difficult to talk to people (or at least, I find it difficult to strike up conversations - after that I'm not too bad). I also feel like my skills aren't very useful broadly speaking. Maybe it's imposter syndrome because I am good at my day to day work (which can be quite complex).
I enjoyed the article though. It gives me hope that perhaps I can push myself and improve to the point where I can make something work. The thought of going out selling is a scary one to me though.