Ask HN: Subscription web sites: How did you get your first customer?

70 points by bkovitz ↗ HN
Question for techies who started a web business with a subscription revenue model: How did you get your first customer?

I'm asking because I'm a typical shy coder: cranking out truckloads of Python code is not a problem, but socializing and having a big network of business contacts is. I'm still grooving on David Heinemeier Hansson's speech at Startup School, and I want to go the subscription route. And I believe that nothing will drive the evolution of a site better than having customers. How to get those early customers, though, looks right now to be a dark mystery. How have other techie folks overcome this problem and gotten those first few customers?

51 comments

[ 3.0 ms ] story [ 76.9 ms ] thread
I don't know you, and I could probably churn out a bunch of semi-valuable tips, but instead of doing that, I'm going to suggest the following: Find a partner that has the business skills you lack in marketing/sales/promotion.

I think you'll be much better off with someone like this worrying about these problems while you handle the technical realities. Once you get customers, the problems of getting them will seem trivial to the problem of keeping them.

Good luck.

We may well bring in a partner who has more sales skills.

However, I'm not looking for tips or advice (right now), I'm just looking for stories of how techie types with subscription web sites actually got their first customer.

Open your product to beta testers, and hopefully once the beta period stops, a certain percentage of the testers will have built a reliance upon your software, and sign up to the paid account.

You can also use the testers for testimonials etc for your site. A referral fee will then help push your current customers to recommend it to their friends.

It'll be a lot easier to find people to sign up for free, then it will to sign them up immediately for a paid account.

You can then convert them later :)

How did you find your first beta tester?
Find an active community website(call it X) that is related to your product and post some messages in the boards offering a free trial to X members. For instance, if your product is for investors, post some messages on Fool.com or MoneyMakersGroup. If your product is for gamers, maybe something on the IGN boards. For hackers, this site would do. For car buffs, maybe CarForums.net.

If it's a B2B offering, find a related vertical and look for forums or industry websites for that vertical. Again, offer a free trial or some kind of benefit to entice these customers to sign up and keep them satisfied while you work out the bugs.

Generally there are very active and loyal members at these sites that are very interested in new tools that would appeal to their niche community.

These sound like fine ideas. Are they how you got your first customer? How did that happen? I'm interested in hearing the story from start to end.
Yes, for a project I'm currently working on. It's targeted toward consumers in a small niche medical community. I Google'd until I found a handful of active forums for this niche. Then I posted to the most active one. I offered a free year's trial to the first beta users. I got 12 from that site. Then after a few days I repeated that process on another forum and got 8 more. I'm working out a lot of bugs but having real users is great. So far they've been active and enthusiastic with feedback. I'm not sure whether this is the right niche for my product, but am improving it a lot with their help. I haven't charged anyone yet(in fact, haven't even setup the payment system yet), and may run into difficulty when I do launch the public product since it seems consumers are very reluctant to pay for things online. But that's why it's great to have beta users from my target market who can help me get the product to a point where it's good enough to charge.
Wow, so you got your initial users by posting on-line! That's interesting. So far, I had been assuming that we would have to work through our networks of friends and/or make cold calls. Cold-calling is probably the hacker's worst nightmare, but I've been revving up: "I can learn to do ANYTHING!!"
From the point of view of a customer/beta-tester:

If your problem solves someone's need chances are they will find you. If posting a link on one or more sites frequented by your target audience doesn't generate a single beta tester you might want to re-evaluate what you're building.

For example I was a beta tester for GitHub and converted to a paying customer after the beta period. Someone must have put an announcement here about the site and since it solved my need for git hosting I immediately signed up.

For me, I get beta testers for a site I'm working on through my blog. I never advertised it (other than a link in my "Projects" page), and admit up front that it isn't even beta-quality yet. Still, I get about 10 signups a month.

Write a blog and publish the articles on social networking sites. You'll get beta signups naturally from that.

I concur on the beta tester conversion path. It's a twist on the freemium model without the negative that people expect some level of your product to always be free. Whenever you're ready to stop giving away from service, you announce the end of the beta. The freeloaders leave without a fuss, the real users pony up without complaint...because you were always up front that it wasn't always going to be free.
I'm sure this is hopelessly old-fashioned, but we advertised. Our web business is aimed specifically at artists, so we advertised in print magazines aimed at artists, handed out fliers at art exhibitions etc.
Thanks! What web business is it?
Nothing very interesting to the high tech sophisticates of Hacker News, I suspect. We're a one stop shop for artists who want their own websites - we'll register their domain name, sort out their hosting, give them access to our CMS so they can create and maintain their website. But the point of my post is that as long as you KNOW WHO YOUR CUSTOMERS ARE, you can be creative and find cheap ways to reach them. It costs us almost nothing to get some postcards printed up and spend a few hours enthusing about our service to artists at a local exhibition. And I'm sure you can do the same, even if your service is less tightly focused. If it's aimed at students and young people, think about legitimate ways (that won't get you thrown off campus) of handing out fliers to students at your local university. Then think about how you enthuse and incentivise your friends and friends of friends to do the same thing at the universities and schools they go to all over the country. If it's aimed at small businesses, write to or call in at one hundred local small businesess. It may sound hopelessly inefficient and low-tech, but it will get you your first customers. And we've found that word of mouth does work (despite what Seth Godin says). The other thing I can't emphasise enough for a subscription service is that it needs to be REALLY easy to sign up for and try out - Paul Graham has written an essay on this I think.
Realise that may not be very helpful if you're shy. That's probably a whole other issue to work on.
I've actually had some success on the shyness front. I did two things here: took a course on improvisational comedy, and hung out with some pick-up artists who showed me how they start conversations with women. These things got me into the mindset of "when something looks like I couldn't possibly do it, test that hypothesis." Last Saturday, I actually got a date by walking up to a woman on the street and asking her! Later that night, I had dinner with her and her family. Even a year ago, this would have been unthinkable.
Well done - that's impressive.
Interesting stuff, and thanks for the reminder about keeping the sign-up process super-simple. We are indeed aiming to make something super-simple. Getting to know the customers is top priority--hence this post.
One other thing I've learnt - think carefully about how technical your users are.

When I started out I knew my users wouldn't necessarily know what HTML is, or CSS, or DNS, or FTP. But I did assume they'd know how to crop a photograph; have some understanding of what a web browser is; know the difference between the browser address bar and the Google search box; and know that you may need to press Refresh to see changes to a page. All those assumptions proved wrong for a significant minority of my customers. Which (at the risk of stating the obvious) isn't their fault for being non-technical. It's my fault for not understanding my customers.

So did it work? Can you tell us more?
My first customer was not a paying customer. It was probably my 4th or 5th that actually paid. The first people are really beta testers.

I knew people liked it when they asked what it cost.

Disclaimer: I'm still a long way from profitable, and I'm not totally committed to the subscription model. In 2007 my subscription revenue was $70.50 and ad revenue was $2.05.

Thanks! Interesting data point that you didn't charge your first few customers.

What is your site?

With respect, you can't call someone a customer if they are not paying you. There is an enormous difference between a user and a customer (or a prospect and a customer). You should reserve the use of "customer" for folks who are actually paying you. Giving a service away doesn't create customers (it may be a part of your customer creation process but they become customers when they pay you). If a "beta customer" does not pay you they are not a customer (yet). There are distinct methodologies and processes for customer development in the same way that there are for product development: one excellent book on Customer Development is "Four Steps to the Epiphany" by Steve Blank, which is chock full of practical advice on developing early customers.
I think it depends on the demographic of your prospects. For some demographics "user" and "customer" aren't so far apart.
If you are selling advertising then your customers are your advertisers, not the audience that you supply them. Users may be customers in larval form, but it is a significant step for a subscription model business to convert a user to a customer. Many startups fail to pay attention to this soon enough (e.g. before they go broke).
Adwords! It's converting profitably for a subscription site I'm helping out with. Can't wait to qualify for Google CPA - http://adwords.google.com/support/bin/answer.py?answer=60150...
Wow, so AdWords has actually worked for a subscription site. Is this how you got your first customer, or how you expanded the business after the initial few customers?

I'm looking for how to get that evolutionary-feedback process started.

We started with AdWords and it worked pretty well. Start researching what ads show up when you search for things that you think might lead to your site to get a sense of what ads you'll be running up against.

We made a big mistake when we launched to not have a free version that a lot of our competitors offer. Our ads would get clicked on, but they'd bounce almost immediately. Once we added the free version, our ad clicks went up substantially, and this time they actually created a free account and the increased volume and retention has led to more sales.

Launch site, no inbound links, nothing indexed in google, buy ads, get subscribers. Now, 6 months later, pages are getting indexed and starting to rank higher for keywords that actually convert. About about 40% of signups now come from adwords the rest organic.

Putting videos on youtube proved to be a quick and dirty way to get visibility in search results while the site was nowhere near the top 500 in search results.

Only problem is it's a small, niche market and it's hard to get much quality traffic. Still, an extra 2-3K a month is always nice to have around.

> an extra 2-3K a month

This is what I am trying to do. Please share any other tips that you have please.

Adwords/AdCenter has worked for me as well so far.

My tips:

1. Use conversion tracking. Dump keywords that don't convert. Try to come up with keyword combos that beat the keywords you currently have.

2. Use negative keywords if needed.

3. Try different ads. Dump ones that don't work and try to come up with ones that can beat those that do.

4. You don't have to be the top bidder to get clicks and conversions. I've done better in terms of conversions and cost/conversion in 5th place than in 1st.

5. In my case, avoiding the "content" network was a great move. My ads only show up in search results, not on parked domains. I refuse to use Yahoo! Search Marketing for the simple reason that they don't let me turn off the so-called content network.

What you're doing is direct marketing, so your ad copy in the ad is important as is the copy on the landing/conversion page. Don't be afraid to do A/B testing to see what copy works best and dump the old copy. Write copy that tries to beat the winner and test it.

Offer a compelling free service that reels users in and then sell added-extras in the subscription account that they can't live without - aka the "freemium" model.

This is what 37Signals do so well, we operate a similar model at Zamzar and more recently Animoto seem to have struck gold with this approach - with them you can cut a 30 second video mix of photos and music, but if you need more time you have to pay.

Seconded. This is what I did with a company I started three years ago (and sold, at a good profit, last year). The audience was there, and they were used to getting similar services for free. By having a few extra killer features and being more innovative generally, 5% of free users converted to paying subscribers.

Sadly this percentage was far too low for it to become a 'big' business and my knowledge of how to grow a company was very poor back then (I used to believe you got the product right, then sold it.. no sirree, that's not how big business really works).

What were the services and new features? And of course, how did you find those first few customers, so you could tell that they really liked the new services and features well enough to pay for them?
Good questions. It was an RSS filtering, reformatting and republishing service, so the audience was mostly geeks and what I call "media geeks".. people who run Web sites who wanted to republish blended feeds on their sidebars, stuff like that. One important unique feature in our case was that NO "Powered by Whatever" nonsense was added into the republished versions people put on their sites. Secondly, we let people blend feeds, rather than just republish a single one (this was very new in 2005!). We also supported filter by query, stuff like that. Oh, and perhaps biggest of all was allowing users to ENTIRELY change the template, free-form. We had predefined templates, but gave the user TOTAL control over the output.

I'm not really sure about how those people found the service. There were a few minor writeups on reasonably unpopular blogs at the time, but 100 signups a day were coming in so I didn't care. Unfortunately I focused on the technology instead of the business so never bothered to keep track of it.

>> (I used to believe you got the product right, then sold it.. no sirree, that's not how big business really works).

What would you have done differently in retrospect?

I'll say what I'd do now, and what I'm about to do on the "next" thing, having gained quite a lot of education and mentoring on the "business" side of things since then.

I'd start selling. Day one. Even if the product or service I'm selling still needs to be developed, I'd work out who wants what I'm developing, what they'll pay for it, and how I can sell to them. Once I have the product working in a saleable form (think 37signals "Getting Real" levels of "done" here - not perfectionism at all), I'd switch 80% of energy to selling it. Cash is king. With cash coming in, I can pay other people to do the grunt work as well as work on improving the product.

But.. selling is absolutely key. Last time I focused on making a kick ass product, and it worked, but I didn't know who the market was, and all the sales were from luck alone. I didn't sell the product whatsoever. It sold itself, but that's not a good strategy unless you have a mass market product. If you have a niche product, you need to be selling it yourself from day one.

I would highly recommend a book called "Ready, Fire, Aim" by Michael Masterson ( http://www.amazon.com/Ready-Fire-Aim-Million-Agora/dp/047018... ). It outlines the whole process of successfully generating a high revenue business from start to finish, and is written by someone who has launched several businesses that have gone on to reach 8 and 9 figure revenues.

I can vouch for the freemium approach. Your free service is there to create word-of-web and word-of-mouth growth. You may even want to hold off on the premium version for a year or so. There's some conflict between being non-commercial and buzzworthy, and pushing your premium version to get subscriptions.

Learning about your customer is crucial. The easy and cheap way to do this is run a survey on your site, while your service is still in the free phase. You can promote this survey aggressively to a portion of your user base and expect around a 15% response rate. One of my businesses had some obvious quantifiable differences between a free and premium version, so the survey included these 3 questions:

* Would you be willing to pay more for an enhanced service?

* How many Xs, Ys and Zs would you want?

* How much would you be willing to pay for this per month?

After collecting thousands of responses, I ran them through a Bayesian clustering algorithm (which can handle missing answers) and it spat out a bunch of service levels and price points that made a lot of intuitive sense. These were $5, $10 and $20/month, and later we added $40/month. Interestingly there was one big cluster that wanted the world for $1/month, whom we naturally decided not to cater for. We also learned that some features, such as RSS, didn't correlate at all with people's willingness to pay, so we added them into the free version.

Net result after a few years is about 1.5% of active users are on one of the premium levels, with the number approximately halving as you go from each price point to the next. Since the price doubles at each level, the total revenue from each price point is about the same. And it's a very tidy living for me and my partner.

One last thing: pay a lot of attention to what happens when people's subscriptions end involuntarily, because their credit card expired or stopped accepting payments. This happens a lot. You want to target these people with reminders about the premium service's benefits, and a discount promotion to sign up again with their new card.

I tell you what we did (which seems to be working):

1) Launch your free product (the free part of the freemium). 2) By the time you get it tuned up, you'll have a strong sense of where the pain is (and hopefully a reasonable guess of where to put the pay wall). 3) Put up a little permission marketing form that essentially says, "Hey-- we're going to launch a premium product soon that does this, this, and this. If you want to hear about the launch, give us your email address". Measure how many of your free customers sign up for that. 4) Look at the resultant list for interesting email addresses and send a personal email to some of them asking them if they'd want to dive in as pilot customers and help shape the product.

#3 helps you confirm your understanding of the value of your premium offering. #4 gets you your first customer.

If you don't have the sales-fu to reach out to pilot customer candidates, just launch it quietly, email a subset of the list who signed up to hear about it, and measure signup percentage (which will be WAY lower than you think it will be).

(comment deleted)
My product is based on a subscription model. Here's what we did. First: Make something that people want and would be willing to pay for. Second: Tell everyone you know that's what your doing. Ask people who they know that might be interested. People are usually willing to help and give you a name or two. Third: Don't count-out old-school marketing. Cold calls, business events, etc, go to them. Meet your competitors, usually they won't be adverse to this and you shouldn't be either, you can always learn something from them.

Know your competitors very well, and know what they do well and what they do poorly (or at least what you do better). This is important.

Third: Have a working demo. I don't give away free trials, I have a demo. This may or may not work for your product.

Fourth: Follow up with everybody, always send thank you notes, and follow up a week later, every week until they give you some straight answer.

I'm a coder who is very new to sales, but I've had some good people around me that were quite good at sales. Meet those people, show them your product, ask for some help.

Ok, to actually fully answer your question: My first customer walked into us. I've been doing websites for years, and I've set up some content-management systems. Most of the time people just don't like them. We made Servee, and took it to our current clients, and moved them over. Our next clients came from this above (cold calls, networking, and sales)
What did you make that people wanted?
We have a hosted content management system at Servee.com Keys: If you're in a market where there are many competitors, separate yourself by doing SOMETHING better (price, features, experience or a combination) Designers like it, so they sell it to their customers; They like that they can give their clients full content control and powerful tools, but not design control. End-Users like it, because we test it on people that have very minimal computer experience. This means it becomes very intuitive to use. Our UX is excellent.
Thanks for the info! Interesting how you found a way to differentiate yourself from competitors.
The teams I know that were able to attract early customers had a sales pitch (a clear benefit, demonstrable difference from alternatives, and a reason to believe) and were "open for business" (had a company and a way to accept payment). It's hard to tell where you are in the cycle but those would be two places to start.
Find the biggest communities dedicated to the niche your software services. Big Boards is a good starting place. http://www.big-boards.com/

Then jump into the forums and start posting responses. You don't want to come off spammy, but definitely try to slip in references to your site whenever applicable.

Are you solving a real problem for someone? Whose problem are you solving? You should have a flesh-and-blood person you can point to (not just "surfers", or "facebook users"). If you don't actually know someone who has that problem, how do you know people are willing to pay for it?

So I'd say you have to know your first customer, and build from there. If you don't personally know them, go to a meeting, conference, or mixer and meet them. As uncomfortable as you may be, just do it, or find a partner who is good at it.

Maybe for a free site like facebook, you can "build it and they will come", but if you want people to pay you, you have to get up from your computer and sell your product.

At the risk of stating the obvious . . . .

Don't build your app and then look for your first customer.

Find your customer first, find a need, and then build an app that fills that need.

Listen carefully to things people you know say, if you're paying attention, you will hear tons of needs waiting to be filled.