Over the weekend I finally completed everything for the SportsLizard revamp I was working on.  Yesterday I made the official announcement on the blog and sent out a newsletter.

Now that I’ve had a few days to calm down from everything, I’m not really sure how I feel about it.  Weird is the best word to describe it.  This wasn’t a “typical” project for me/us.  Usually there’s the hope that a project will hit it big and make us a ton of money.  While I’ve always thought this had the potential to be huge, that wasn’t really what I set out to do.

First and foremost, I just wanted to better capitalize on the traffic we were getting.  That’s why I put everything else aside and took on this job.  I thought the old site did a crappy job of that – I felt like we were leaving free money on the table.  Immediately it looks like my fixes worked:  it’s been over a week since the site got up (albeit not complete until the last few days) and right away page views and time spent on the site essentially doubled per visitor.  We cleaned up our ads, getting rid of AdSense and using almost exclusively higher paying and more relevant eBay ads.  Ad clicks are way up too.  So in this sense it’s been a success.

My second goal was to learn.  SportsLizard has always been my playground for learning how to design and develop a site.  It was my first site, and ever since then there’s been a major update every year or two.  Each time I’m trying to push the boundaries a bit more.  In doing so, I’ve taken my programming skills a bit further.  On this revamp, I’m most pleased with my improvements as a designer.  My graphics skills and CSS skills are a lot better than they were a few months ago.  I upped my PHP skills as well.  I’m really happy with how the new-and-improved Price Guide came out.  So again, in this sense the project has been a success.

My third and final goal was to provide something more useful to the card collecting community.  This is a tough one to evaluate.  The really hardcore community, the collectors that spend time and money in the hobby every day, the ones that could make SL HUGE, have not fully embraced it.  I can hypothesize why, but if I knew for sure I’d make the changes necessary.  On an individual level, I can get influential bloggers behind me (they already kind of are).  But the guy who spends all day on the card message boards seems to have no use for what we provide, even the Price Guide tool.  This makes me scratch my head and quite frankly pisses me off.  Part of me thinks it will just never happen with SL – we have our audience, our niche, and I should leave it at that.  Part of me wants to keep pushing and make it finally pull through.  Another part of me wants to take everything I learned with SL and start another unrelated cards site and focus on that (a side project I’ve been toying with for a while, one that I think the hardcore collectors would eat up, which would bring me my vindication in a totally different way).

The project also took a whole lot longer to finish than I anticipated, which kind of pisses me off because I feel “behind” even if I’m really not.

It pretty much comes down to this:  I badly want to see it become a bigger success, so it’s tempting to try to make SL my entire focus for a while. But that would be a bad move.  From a strictly business standpoint, SportsLizard accounts for a very small part of Pure Adapt  – it has no expenses, so the revenue is all profit, but that’s still only about 10% of our profit as a company.  E-commerce is what drives us and will continue to drive us.  I’m a firm believer in all of the benefits of working on side projects like SportsLizard and Music-Alerts, but they are just that:  side projects.  Until  some other project forces us to divert attention away from e-commerce I am going to try to get back to limiting myself to 5-10 hours a week max on those types of projects.  They’re great learning experiences, they make a little cash, they bring some pub to our company, but at the end of the day they can be a distraction if you don’t cap your time and return focus to more important tasks.

I decided to take a wholly fresh approach to the design and programming.  I wanted to take the extra time to learn and implement new CSS and design skills, and also to efficiently organize all of the code.  Aside from a relatively tiny site like Music Alerts, this will be the best overall job I’ve done with the entire site development process.

In doing so, I got to extensively look over a lot of the PHP code I wrote in April 2007 when the price guide portion of the site was launched.  It really is truly amazing what a year and a half will do to your skills.  Every file I open I think to myself “why did I do it this way?”.  While adding features, I’ve proceeded to write cleaner, more organized code.  Many pages have gone from 1,000 lines of code down to a few hundred while increasing functionality.   At some point in the last year or so, there was a shift in my development skills from simply being able to do almost anything, to being able to do almost anything in the most efficient way.  When I say most efficient, I mean less lines of code and more organization, while factoring in user interface, SEO, database design, and scalability issues.

If you constantly attack learning with a passion and are never satisfied with where you’re at, you’ll be stunned at your growth.  Just like we micro innovate as a company, you can improve your skills dramatically by getting a little bit better each and every day.  In our “want it now” society that’s not really a sexy proposition, but it works.

When SL rolls out in a few weeks it’ll be the first complete example of a site I developed at this level of programming.  And while this could bring in some great supplemental income for us, the true benefit of these skills will come in 2009 when Mike and I tear apart the shopping cart platform we built for Detailed Image and Tastefully Driven and rebuild it from scratch.  It’s good now, but we want better code, better organization, better UI (including useful AJAX), and a better back end.

The benefits of such a platform for our company will be huge.  It’s not just about increased sales or better shipping efficiency.  It sets the system up to be completely scalable.  The shipping system will be so easy to understand that any warehouse employee can fulfill orders successfully.  The inventory system will be so well organized that our first warehouse manager will be able to jump right in and manage our 30+ vendors.  The programming will be so well organized and documented, that any good programmer will be able to understand the structure immediately.  So by 2010 when we need two or three dedicated programmers, we can hand them this beautifully developed cart that they can dive right into and grow and expand.  If one programming issue impacts both sites, they will only need to change the code once to deploy it onto both sites.

All of this leads me to the (somewhat long-winded) point of this post:  if you really want to innovate on the web, if you really want to bootstrap an online company, you need to learn how to program.  Pick up an introductory PHP/MySQL book.  Learn HTML and CSS.  Get a solid understanding of Javascript, XML, and eventually AJAX.  Learn the basics of running a server.

This didn’t happen to me overnight.  Most of it – especially the server stuff – happened because of need and not because of desire.  But every single bit of it was worth it and I’m glad I learned it.  Because it gives me power.   Any website that I can dream up, I can create with $10 for the domain plus my time and the space on our server.  How cool is that?  I don’t need to search for a programmer.  I don’t need to try to convey my ideas to someone else and have things get lost in translation.  I can just do it and see the vision unfold.   Every idea I’ve had that I really wanted to do – SportsLizard, iPrioritize, Music-Alerts – I’ve done.  How many people can say that?

And even if you eventually end up passing on your programming to employees as we plan to do, it’s still a huge business advantage.  You can communicate better with your programmers and other managers because you truly know how things work, what can and cannot be programmed, and the difficulty involved in getting a feature programmed and deployed.  Understanding your technology inside and out helps you with your marketing efforts, and it helps you with your customer service.

When it comes to web business, programming skills really do equal power.  So young entrepreneur, instead of writing long winded business plans or reading more generic business books, I suggest forking over $100 on programming books and some web hosting.   Trust me, it will be one investment you won’t regret.

Before getting into exactly what we did and how, I want to preface everything with:  I do not recommend implementing this for a brand new site/business.  In the beginning it is important – especially as an owner – to have contact with as many potential customers as possible.  Your initial customers will likely give you important feedback about important adjustments you need to make to your business.  If you make it hard for them to contact you, I think you’re impeding your chances for success.

But after years and years of customer service, as is the case with both Detailed Image and SportsLizard, the same questions keep coming up and you just end up wasting your time copying and pasting the same response over and over.  For both your sake and the customers sake, making those answers readily available is a smart decision.

The Problems

We had three types of customer service contacts that were repetitive and time consuming:

1. Detailed Image emails – questions about what products to buy, where we ship to, shipping quotes, technical site questions, etc
2. Detailed Image phone calls - very time consuming.  The same questions as above, but dragged out over the course of 30 minutes or more many times, especially when discussing their cars and specifically how to detail them.  Now, these longer calls also generally resulted in large sales, so there was some risk involved.  However, in addition to the time spent on the phone we would also have to manually put the order through the website, which added an extra 5 – 15 minutes.  We knew it was worth it to take a stab at significantly reducing these.
3. SportsLizard emails – mostly about the Price Guide:  how to sign up, how to cancel, and how to get better prices.  Also, people emailing in customs for the gallery was very time consuming for me to post.

Our Solutions

There are more complex solutions to these problems than what we came up with.  We were trying to get the most bang for our buck.  No sense in spending a week programming a solution when you can spend an hour and get the same results.  If the results fell short, we probably would’ve gone back to the old contact pages and spent time developing a more complex question/answer system.

Detailed Image:

Greg put together a FAQ of the most common questions.  I turned it into a “drill down tree” FAQ to replace the contact page.  So instead of a contact form that included our email address and phone number, customers are presented with the following:

Which drills down like so:

At the end of certain “branches” we give our phone number and/or email address, only when necessary.

SportsLizard:

Replacing the contact form with a page that has the top 5 most common FAQs, a link to a regular FAQ page for the rest, and the text “I checked the Frequently Asked Questions and they did not answer my question.”, which then reveals the contact form.

To minimize the time it takes to post a custom, I created a custom submissions form.

The Results

Detailed Image:

For the email side of things, Greg reports a huge drop off in basic questions like “how do I log in to the site?” or “how do I get a shipping quote?”  We are attempting to quantify the improvement, but the data set is incomplete at this time (it’s hard to determine what exactly constitutes a “contact” via email so it takes quite a bit of work to comb back through emails and only count specific instances).   That said, contacts are definitely down.

The phone is where we’ve seen a HUGE improvement.  I went back and studied our bills for the past few months. In March we had 36 incoming calls and we checked our voicemail 16 voicemail times.  In June we had 2 incoming calls and checked our voicemail 15 times.  That’s a 94% reduction in incoming calls.  We went from handling almost two per business day to handling none.  That’s a lot of time saved.

It should be noted that an incoming call only happens when the phone is left on, and we’ve been turning our phone off with the exception of turning it on to check for voicemail occasionally.  Keeping the phone off has the added dual benefit of allowing us to stay focused without distractions, similar to keeping your inbox closed or IM off (for another post).  So the combination of  adding a FAQ section, making it harder to find our phone number, keeping the phone turned off, and steering customers to email via our voicemail message has essentially eliminating incoming calls.  On the rare occasion that we do get a voicemail, I’d say half of them have already been resolved by email or by the customer themself and don’t require a call back.

SportsLizard:

I haven’t tallied the data, but I’d estimate I went from about 5 customer service emails/day to about 2 emails/week, a huge improvement.  I really love making people click the link that says “I checked the Frequently Asked Questions and they did not answer my question.”  That extra step of forcing people to click the link makes them spend a minute really thinking about their problem and checking the FAQs before jumping in and contacting us.  I really like this solution and will definitely use it in the future on other sites.

On the customs side of things, I probably went from spending 2-3 hours a week posting in the gallery, to spending about 10 minutes a week approving submissions in the database.

Financially:

Detailed Image is continuing to exceed our expectations, with 100%+ growth each month compared to the same month in 2007. SportsLizard – despite my minimal effort – has seen an increase in paid subscribers to the Price Guide and an increase in ad revenue.

Customer Satisfaction:

No, we haven’t done a customer satisfaction survey, but here’s my opinion:  most customers don’t notice a difference, some customers are more satisfied because they get their answer immediately via FAQs or our detailing guides or other info on the sites, and a small minority are mildly inconvenienced because they want a person on the phone right now.  For some businesses that aren’t e-commerce, that minority would be a majority and a system like this wouldn’t make sense.

In our case, we’ve made a commitment to minimize unnecessary contacts and funnel the necessary ones to email.  We have received a few emails from people who complain that they can’t find a phone number.  In these cases, we either apologize and answer via email or pick up the phone and call if the situation warrants it.  The difference of course being that we’re the ones deciding whether or not a phone call is necessary.

While I’m sure we’ve lost a sale or two because it’s harder to find our contact information, we’ve more than made up for that in revenue generated from the time saved.   It’s OK not to appease everyone – we know that the majority of our customers are happy and praise our sites for having exceptional customer service.  We’re now able to get to the important customer service emails faster.   I personally will sacrifice the profit from one pain-in-the-ass, hold-my-hand customer for the free time to pursue customers that value our time and won’t bother us with unecessary emails/calls without first checking the site.

The bottom line:  it seems as if my hypothesis – that putting a customer service system in place would save time without sacrificing revenue or customer satisfaction – was a correct one.   

However, this morning my biggest SL nightmare became reality – the Price Guide software mysteriously stopped working.  A bunch of cancellations followed, as did an email from a loyal user who has priced 11,238 items using our Price Guide (yup – he’s run 11,238 searches!).

This was at about 8:30 AM while I was sitting in a Starbucks.  I had planned to do some relaxing work and then meet my cousin for lunch…but all of a sudden I was in panic mode.  The application relies on a lot of things to be working right, most notably our syncing with Google Base, and I had no clue where to start.

I thought Google might have disabled our API keys…nope.

I thought someone might have hacked the SL database and screwed with the integration of SL and Base results….nope.

I thought that eBay, NAXCOM, and Beckett could have all stopped listing items on Google Base (essentially killing the service)….nope.

Then I tried a pre-launch beta version that I made that spits out results as a text file without any graphs or a fancy UI.  Amazingly that worked.  So I went line-by-line through my code and compared the beta version  to the live version.  After about an hour I figured out that a filter I had put in to weed out items returned with a $0 price was no longer supported (correctly at least) by Google Base and was throwing the results off.  The same search would go from 30 legit results to 500 with the filter removed.  Very rarely are items listed at $0 anyway – I just put that in there as a redundancy – so I removed it and everything began to work fine again.

The last time I successfully used the Price Guide was a week ago, so it looks like it was in bad shape anywhere from 36 hours to 1 week.

Now the hard part – damage control:

• At 8:30 I had immediately placed a notification in red on the Price Guide page saying something to the effect of “We’re experiencing technical difficulties, we’re working to resolve them ASAP, we apologize for the inconvenience”.   I took that down once I was certain I had solved the problem.
• I emailed that uber-loyal user back who initially alerted me of the error. I refunded his most recent payment and offered him a free premium account for life.
• I emailed everyone who had canceled in the past week apologizing for the inaccurate results and offering them the same free premium account for life.
• I wrote a post on the SL blog (which is auto-emailed to newsletter subscribers as well) notifying everyone that the issue had been resolved and extending the same free premium account for life offer to anyone who I did not email that also experienced troubles with the Price Guide during the past week.

Whew.  That was a stressful morning.

As much as it sucked, it allowed me to re-familiarize myself with the app I built almost a year ago.  More importantly, it created an opportunity for me to show my customers how much I care about them and their experience with the Price Guide.   It also gave me a chance to show that we address our issues rapidly – within hours of being notified of them…even on a Saturday – and with a candid, honest, and genuinely remorseful attitude.

No one wants to have technical snafu’s like this, but when presented with one you’ve got to make the best of a bad situation.

Our SportsLizard Price Guide is just about to pass the 20,000 registered user mark. For those who don’t know, a free account gives users 3 price searches per day, and for $5/month you can perform unlimited searches and access a few other premium services. 3 searches per day isn’t much – more of a trial than anything you could routinely use if you’re a collector. Truth be told, we only want people using the site that are willing to pay the $5/month.

To date (since May) there have been ~200,000 price searches performed, mostly by Premium Account holders. So why wouldn’t we open the whole thing up and make it free for everyone? We might be approaching 2 million price searches right now, and with that we could sell a lot of advertising. A lot of people think that way (at times I have), but there are a few fundamental flaws with that thinking:

1. It’s not that easy to sell the advertising. Ad networks are OK, but if you want to make more than a few dollars CPM (cost per thousand impressions) you’ll need to sell ads directly. I’ve negotiated quite a bit for SportsLizard with the “best” companies in our industry – many of which have contacted us – and it’s REALLY hard to get that from them. They want cheap ads ($1-$2 CPM) because some other stupid site owner will offer it to them. Never mind that the other site is plastered with ads and the quality of visitor is low and won’t convert to sales for them – they don’t care/aren’t smart enough to draw that conclusion. They just want to buy cheap ads. Which leaves you with two options: cover your site with a crazy amount of ads or stick with ad networks (AdSense, YPN, Value Click, Commission Junction) and earn a few bucks CPM. Think about that – for every THOUSAND impressions of your ads you’re only making $1-$3. That takes a crapload of scaling to become profitable.
2. If you DO scale that to profitability you’ll be massive enough to start dictating higher rates from people – maybe as much as $10 CPM. But scaling to that size means you’ll need millions and millions of visitors each month. And that means you’ll need multiple servers and a larger customer service staff and more programmers…all so you can make money selling ads. That probably requires venture capital or angel investing to scale the company, and you still aren’t guaranteed anything. Getting that kind of traffic is tough.
3. The goal of your site becomes to get people to click away to another site! To keep your advertisers happy, you’ll need to show a high CTR (click through rate) and, if they have a good analytics program, show them that those customers are profitable for them. I don’t ever want to run a site where I encourage people to leave. This is where online ads differ from newspapers and magazines. If you go to your computer and type in the URL you see in a magazine ad, you still have the physical copy in your hand. You’ll probably keep reading at some point. The content still matters. On a website that baits people into clicking, good content can be a hindrance.

Or we could just charge $5/month/user It doesn’t drain our single server. It doesn’t overwhelm us with customer service issues. At 1,000 paying users, which we should reach by the end of the year, we’ll be making $5k/month in recurring revenue. To make that in ad revenue – assuming you get $5 CPM…which is tough – you’d need to generate 1 MILLION ad impressions each month, not to mention handle all sorts of scaling issues.

Something to consider when you decide to start the next big thing. At Pure Adapt, we’ll certainly take our “home run swings” with cool projects that *could* make a lot of money if they blew up, but our core projects are always going to be focused around things that can generate money immediately without extreme scaling and dependence on advertising revenue.

Now comes the questions: why did PayPal wait several days before acknowledging the problem? Why wasn’t their customer service team fully aware of the issue? Why did they wait until yesterday (4 days late) to post an announcement on their blog? Why wasn’t there an announcement on the homepage so everyone could immediately be aware of the issue? And why didn’t they send an email out to all of the people the problem impacted?

For SportsLizard, it was no big deal. An inconvenience – yes. But we got our money. However, there are several people I’ve talked with that rely on payments and had their cash flow hurt significantly (a few hundred $4.99 subscriptions is nothing compared to people with a few thousand $99.99 subscriptions). I’ve also seen comments from people (these ones from the PayPal blog) that have had subscriptions cancelled by PayPal without their consent:

Ummm, it’s worse than that. You are now cancelling my subscriptions. As far as I know this is not reverseable. Your system isn’t collecting the money for a subscription and now it is cancelling the subscriptions. I have had 12 cancellation just this morning. That’s more than I usually have in a couple weeks.

And…

My web hosting firm has been unable to collect our subscription payments since yesterday. There’s a lot of angry customers too because they’re reporting their subscriptions have appeared to have been cancelled.

And…

I as well have had 5 cancelations over the period of Sep 1st 2007 – Sep 2nd 2007. 1 of these customers has messaged me back, asking why I have canceled the subscription. I pointed him to this website, and he said in his own words.
“Ah that figures, paypal at it again”
Good job guys…

Oooof.  You better fix that PayPal or risk getting slapped with a class action lawsuit.

On a much more positive note, the Price Guide passed the 15,000 registered user mark over the weekend! This is a huge accomplishment for us. We launched May 3, so it has only been live for 4 months…an average of nearly 4k/month! The first month was slow, so if we keep up that 4k/month rate that it’s been going at since then we’d sign up 48k users in the first year! I think once you start talking about having 100k+ registered users for something you enter a whole new stratosphere. On top of that, traffic and ad revenue has more than doubled each month, and again I feel like the site is pushing the limits of a “small” site and has the potential within the next year to become a very very profitable mid-size site.

Now comes the big launch of the Detailed Image site we’ve been working on since June.  That comes this week, and I can’t wait to get that monster up and running.  As always there will be a few hiccups, but once it’s on cruise control our company will really be operating at close to 100% efficiency, something we’ve all been pushing hard for.