Large Language Model ChatGPT has received an upgrade in the form of voice and image capabilities. OpenAI confirms that the language model offers a…
Social discovery app Real Time Wine recently called it quits — a bold move that founder Andy Hadfield credits to “revenue not scaling quickly enough.”
We all know the famous startup maths: nine of out 10 tech startups fail (author unknown), but Hadfield has taken solace in shutting Real Time Wine down, and has written a “lessons learned” post below. It’s a long read, but full of valuable advice for someone on their startup journey.
My Street MBA: Lessons learned from Real Time Wine
Real Time Wine was a retail discovery app. It started in wine, moved to beer and ultimately wanted to trap consumption data and build communities across multiple retail verticals. In doing that, we hoped to personalize and improve the ecommerce experience — one that is so largely dependent on spammy, irrelevant daily emails. I shut Real Time Wine down in April 2014. Revenue didn’t scale quickly enough.
“Lessons Learned” posts can suffer from two fatal flaws. Firstly, they can be schmaltzy. Secondly, they can be preachy. But if done right, they can hopefully give a little back to a community that provided an enormous amount of support during this startup journey. These posts can also offer some form of catharsis for the Founder — which is important. And I wrote a “long letter”, not because I didn’t have time to tell a “short story” — but because this isn’t. I’m going to try and do this right. Chris Poole (4CHAN / DrawQuest) did it right. I’ll take some inspiration from his approach.
“They” always tell you that nine out of ten tech startups fail. So you realistically should be prepared in case you’re one of the nine. It’s still hard when it happens. What “they” don’t tell you though, which they should… Is the enormous amount of experience the startup journey provides. You come out a bit sad, but also a bit rougher, tougher and a much more mature businessperson than when you went in.
Here are some big lessons I learned during the Real Time Wine & MyBEER startup journey. For context, here’s a link to the closure announcement (and the announcement that even though the app and business is being shut down, the community will live on in a combined passion project called Incogvino).
I say “we”, but I really mean “I”. The Curse of the Single Founder.
I was blessed to have an amazing support team around this startup. Active, engaged and responsive investors in Michael Jordaan & Mike Ratcliffe and an advisory network including Josh Adler (Founder Everlytic, Director of Entrepreneurship @ African Leadership Academy), Brett Commaille (AngelHub Ventures) and Keet van Zyl (Knife Capital) among others.
But I didn’t have a co-founder. And I will never, ever, do another startup without one.
In 2012, this idea and passion for a retail discovery app started to form. I laid down the first PowerPoint slides, sketched the first notes and overestimated my first market size. All the good stuff.
At that stage I had a choice, to sit around and wait for a co-founder (something I’ve battled with my whole life) or just HTFU and get going. I chose the get going route, of course. And I have no regrets about it.
But the life of a single founder is tough and lonely. Amazing businesses are built with people, not with products. The chemistry between founders and early staff is probably the single most critical factor in its success. When things are going well, you need people around you to high five. When things are going badly, you need people to prop you up and help you get up off the floor and keep punching.
I would have moved faster with a co-founder/team. I would have failed faster at our first four business models with a co-founder/team. And those are both important factors.
Do everything you can to find that support network. Do everything you can to find a co-captain. Whether that means events like Startup Weekend, Incubators, Speed Dating (cough, nerd!) or general Networking events. Cast the net wide.
On the other side of this argument, just remember that divorcing a business partner is harder than divorcing your wife. Choose well. Look up the word “vesting” — it might just save your life! Nothing is easy, is it?
Pay yourself up front. Burn the ships!
I’m very divided on this one. I didn’t pay myself a cent throughout the whole Real Time Wine story. It was probably the wrong move.
On one hand, I think this showed my investors how committed I was, how I’d go to the mat and do anything to give the startup as much runway as possible. On the other hand, it gave a completely unrealistic view of what the startup would actually cost to run. And even worse, almost distracted me from hiring a team earlier (or at all).
At one investor pitch, I’ll never forget a clever chap saying “Andy, if you’re chasing pennies the whole time, you’re never going to get anything done — you’re never going to go for the big picture.” He was right.
I also remember a talk that a friend (and advisor to Real Time Wine) Josh Adler gave many moons ago — when we were both young and ready to conquer the world. He spoke about the idea of a “one foot in, one foot out” approach — one of the methods he used to start his first company. Keep a consulting gig on the side and a startup on the other side. Keep some money coming in, work on your dream in the evenings and the gaps. It was something I took to heart with Real Time Wine — and probably the only reason we lasted as long as we did.
While the “one foot in, one foot out” approach is still the only one possible for many — you run the very serious risk of distraction and lack of focus. Another friend, Gareth Knight (founder of Tech4Africa and Daddy to many more proud startup failures than myself) advocates the opposite approach. Although to be fair, Gareth tends to raise money on shores slightly more lucrative than SA… His approach is “burn the ships”. Get rid of every distraction possible, so the only positive outcome can be success. Others call this giving yourself “the fear” — the same approach to quitting the job you hate, even though you don’t have a new one.
I guess I find myself looking back and wishing I’d burnt the ships a little more. But I also understand that life stage plays an enormous role. You’ve got a family to support and a bond to pay. But you also don’t want to miss the opportunity to put yourself out there and try something, or even worse, not give it a good enough shot.
So pay yourself up front. Even if it’s R5000 per month. If you’ve raised even a little bit of funding (or have some savings), this number is actually meaningless in the grander scheme of things — but meaningful in keeping yourself motivated. You can always escalate your salary quite quickly — investors understand you have to live as well as work.
Use the “one foot in, one foot out” approach sparingly. It may be necessary, but the real trick is understand when to “burn the ships” and go for it. I never quite made that call, and probably had a bit of a slow death because of it.
To Outsource or Not. How to Develop if you’re not a Developer.
With no technical co-founder, the first major challenge was how to get the product launched and iterated — a classic “rock and a hard place” challenge. If you want to hire a developer or development team, you’re in for a cost surprise — which can considerably shorten your startup’s runway (the number of months you have left before revenue needs to cover costs). After all, developers are the inventors and engineers of our century — they’re rather in demand, and have unfortunately figured that out.
Here’s my napkin maths:
For a mobile play, you don’t often get multi-skilled developers. So you’re going to need one iPhone developer and one Android developer that can hopefully double up as backend developers. Hard core coders like this don’t usually design. So you’re going to need a designer. Let’s say you get extremely lucky and you can grab these guys for R30k/m each (almost impossible in JHB, possible in Cape Town if you get younger guys/gals). Then you pay yourself R30k/m. Assign R30k/m for marketing and ops expenses (which is as lean as it comes). TOTAL: R150k/m burn rate.
If most Angel funding in SA is between the R500 000 and R1m, your runway is, shall we say, limited. 3-4 months to develop a product, launch, test traction, get revenue and cover costs. You can play the numbers any way you like, they usually work out in the same range.
So get a technical co-founder. That’s the simple answer. There’s a reason hustler/hacker teams are so popular in the US — they keep costs down and promote speed. In fact, what we’re seeing more and more of in the US is hustler+designer/hacker teams.
If the stars don’t align, you’re going to be forced to outsource. This was really the only viable option available to me two years ago. I examined three potential sub-options within this choice, largely coated with my own experience.
Outsource to a small dev house/agency. Be careful. Money talks, equity walks in this segment of the game. You don’t want to base your startup on a dev house that will easily de-prioritize you in the face of paying work. Increased risk for decreased cost.
To a big dev house/agency. They get expensive. Ludicrously expensive actually, but that’s where all the good creative and engineering talent is hiding, so they’re entitled to charge what they want.
My general rule is if you’ve got the money to hire a big agency, you should rather hire your own team — it’s better in the long run. A big agency does provide a positive spin on risk though, I’ve often caught myself saying the following when telling stories of my development experience:
“With a small agency or your own team, you’ll pay less, but you’re never really guaranteed an end product. With a big agency, you’ll pay more and you’ll be guaranteed an end product — you just won’t be able to guarantee when you get it!”
To India: you hear some good stories and some horror stories. In my opinion, the only work that can be freelanced out virtually is something so simple you’re not going to get anything lost in translation. And stuff will get lost. Be prepared to rewrite everything for you next version. But if it can get you to a simple prototype — and you can’t, go for it. It certainly is affordable.
I was lucky (and unlucky) as you’ll see. Through some trusted industry relationships, I was able to get a big agency at mid agency rates. It was affordable (in comparison to hiring a team), I was guaranteed an end product. I had no idea when it would get finished.
Managing Digital Agencies and Dev Shops
After a bit of a search I decided to develop my prototype with Prezence, one of the hot mobile development shops at the time in South Africa. It’s a funny tale. Throughout a more than two year relationship with Prezence, I had both the best agency experience I’ve ever come across… And the worst.
For Phase One, the prototype, I cannot overstate how good Prezence was. As a creative team that worked alongside the entrepreneur through all elements of the project, they were amazing: conceptualisation, architecture, wire framing, design, UX, development, testing and launch.
Regarding further development and product work, the creation of MyBEER (Real Time Wine for Beer, kindly sponsored by the forward thinking team at SAB) went pretty well, especially considering some enormously tight deadlines. The only negative was that it probably created more bugs across both platforms than it should have, which then provides an interesting observation. Everyone talks about building a platform — but building a platform is enormously difficult. You have to cater for an insane amount of variables, most of which you don’t figure out till you launch. The ability to iterate is essential.
For Phase Two of Real Time Wine, the idea was to iterate on the code base, implement revenue streams and fix bugs. Phase Two was a complete disaster. There are many reasons for it, some on my side, some on theirs. It was painfully slow, we suffered from large amounts of staff churn with senior developers/architects coming and going — and it literally took more than a year to get through the bug lists.
It led me to a fundamental belief, one that may cause ire amongst my colleagues (and friends) in the agency world. It is a belief cemented by the Real Time Wine experience, but one that I’ve seen occur time and time again across a 15 year career: I don’t think digital agencies are structured or good at iterating on high risk, high speed consumer products. Simple as that.
I think digital agencies are at their best when there is a beginning, middle and end. It fits their scheduling better, it avoids complications of staff churn and prioritization. Speed isn’t as essential. They’re just better at once-off builds, be it campaigns or products. If iteration is required, it’s usually a slow, methodical process with a big corporate client that values consistency and lowered risk over speed-to-market.
So, if you’re outsourcing, go into it with both eyes open. An agency or a dev shop will never be your team. They just can’t fulfill that role. I probably got too close to Prezence in Phase One of the project, such was our success. I began to think of them as part of the team — and while some of them would have loved that, the reality is that a startup is always one small, highly demanding agency client in a portfolio of many.
Use external suppliers for what they’re good at — and have a different plan for reaching and achieving scale.
The importance of channel focus in lean startups
One of the biggest lessons I learned was around focus. It appeared in two fairly costly instances.
Firstly, in early 2012 when we were planning the build, BlackBerry still utterly dominated the South African smartphone landscape. If you remember back then (it wasn’t so long ago, but it feels like it sometimes) South Africa had quite a fragmented smartphone penetration. It still is in many ways. There was definitely no cut and dried approach towards which platforms should be supported and which shouldn’t.
Using the app wrapper approach, we could release on iPhone and Android as apps, then still be available to Blackberry users via the mobile browser. We could deploy quicker (Apple App Store had a one to four week approval time back then) and as an added benefit, we wouldn’t have to build a website — we could just render the mobisite inside a picture of an iPhone. We patted ourselves on our collective backs, thinking we’d been really clever.
Turns out, Blackberry users before BB10 came out weren’t app users. Turns out Android was horribly fragmented and the app wrapper approached just garnered frustration from users used to US/Europe quality native apps. Turns out the amount of time and cost involved in getting people to discover you inside each app store actually slows growth, not promotes it. Turns out there’s a whole branch of SEO related to app store discovery that requires an immense amount of effort. And turns out that app users are fickle, a couple of one star reviews can really sink an app in rankings — and users wield the power these days. If you’re not a five star app, you inherit a nasty line item to your advertising budget: help people find this app in the first place.
There’s a reason US/Europe startups usually launch with iPhone first. Channel focus allows you to put all your energy into creating the best experience for your users. iPhone users tend to be app early adopters (although this is slowly changing as Android becomes more mainstream). I honestly think we would have done just as well and got just as far if we’d only released on iPhone. Focus is your friend.
Focus reared its head again with Phase Two of Real Time Wine development (roughly costed at half the funding raised). The idea was to iterate off the base, improve the product and add features that would help us generate revenue. We added activity streams, the ability to attach price specials to locations, the ability to deliver content and ads to people as they reviewed a wine (at the “moment of consumption” — still the cleverest thing I think I came up with).
Phase Two was a mistake and it should never have been built, for two important reasons.
Firstly, your prototype should have some way to get you to revenue, even if it’s extremely simple. Mine didn’t. The further you push revenue down the timeline, the less likely you are to actually test your business. It’s a cardinal sin and one I won’t make again.
Secondly, Phase Two took six months of focus away from grinding out revenue and traction. It was a wonderful and ultimately useless distraction.
It’s easy as an entrepreneur to get distracted by the cool stuff. Designing products is way cool. Grinding out revenue is not. I got distracted. I should have put that Phase Two money straight into marketing and staff and scaled a “team” much earlier.
Don’t believe your PR (too much)
Real Time Wine won plenty of awards and due to the kind community of journalists who supported us, it had an enviable amount of coverage. If I’m honest about it, I think I got a little distracted by the product’s media success. When the world is telling you (mostly) what a great product you have, it becomes a little easier to focus on making that product and its associated community sexier — instead of grinding your way to those user- and revenue milestones.
In a small market like South Africa, especially with Silicon Valley-style products (those that rely on sex, sizzle and scale) — you can’t get distracted, we just don’t get the same kind of runway that Valley startups do. That is both frustrating and liberating.
PR helps you craft your story. It doesn’t help you execute and operate. Lots of PR doesn’t mean you’re winning. Lots of revenue does.
Actionable Metrics: The Wonderful Art of not Deluding Yourself
The journey of entrepreneurship is about constantly avoiding your inherent ability to delude yourself. That may sound funny, but it’s true. Humans are very good at post-event justification and in a startup, that’s dangerous.
The Lean Startup talks about the difference between Vanity Metrics and Actionable Metrics. This is possibly the single most important lesson you need to take into a business. Any business.
There was a beautiful (and embarrassing) moment after I read that book, looked at my management reports that I’d be sending and obsessing over for the first year and half of the startup — and swiftly realising they were all vanity metrics. Downloads. Registered Users. Unique Users. Total Reviews. Total Ratings.
Vanity Metrics are great for the warm and fuzzy. They’re usually cumulative and therefore pretty much always go up and towards the right. They make you feel proud, they pump you up — and they’re pretty much useless for making decisions about your business.
Actionable Metrics are the tough ones to look at. I’m not going to repeat what The Lean Startup teaches — but when you start looking at Revenue Per User, Active Users, Referral Mechanics, Conversion Rates, Cohort Segments. That’s when the truth hits you hard.
Active Users was a particular eye opener for me. It’s an important metric for both consumer products (like the app) and ecommerce portals — an indication of a user’s likelihood to do what you want them to do: submit a review, buy a product etc.
Being honest about your Active Users (and how they’re actually active) will also remind you of a lesson further up in this epistle, the importance of scale. Social systems tend to work on a ration of 1 / 9 / 90. 90% of people consume content or just watch what’s going on. 9% contribute in a small way (click Like or vote for instance) and only 1% actually create content. Some social systems do better at these ratios (like Facebook with it’s frictionless sharing approach) but most tend to toe this line if you really dig into it.
I thought we were pretty special with a 10% create/contribution rate (I defined Active Users as anyone who was earning points on the app — most likely through the creation of content). In reality though, 10% of 4000 is only 400. Not a big number at all.
What’s worse was that the contribution rate was pretty flat through the lifecycle of the product, indicating we got new people to jump on try it out. And then kind of like FourSquare, they got bored and moved on. We had user churn that wasn’t saved by user adoption.
If I had realised this a bit earlier, perhaps things might have been different. Go read that chapter of The Lean Startup. Put together some Actionable Metrics — and more importantly, action them!
Feel free to gaze lovingly at your Vanity Metrics, but don’t get distracted by them.
Fail fast. Get to revenue as quickly as humanly possible.
It sounds silly, doesn’t it? Like something every entrepreneur should know. Fail fast: the ultimate cliché. But you’d be amazed how easy it is to sidetrack the importance of revenue in the Silicon Valley “business model comes second” world we live in.
I only read The Lean Startup two years into the Real Time Wine journey. That was a great pity. Amongst other clever things the book covers — it preaches the fact that you need to get to revenue before anything else. R1 in the bank account is the beginning of you validating your assumptions, testing your products and proving your business model. We failed too slowly to test our revenue models, partly because our original plan relied on the enterprise sales cycle and partly because we didn’t build a revenue model into Phase One of development.
Failing fast, if you can do it properly, is absolutely liberating — because it leaves you with enough time to pivot. Failing slowly is like cancer. You know in your heart that it’s inevitable, but you hang on in order to “tick the boxes”. By the time the boxes have been ticked, you’re out of runway.
I had four business models at Real Time Wine that gloriously failed. It’s interesting to look back at them and try understand why they failed. If you can figure this out quickly, you can build a model that works. That’s what a startup is all about.
BUSINESS MODEL ONE. Real Time Wine was never about wine. Funny that. Wine was a test bed for a much broader idea. I wanted to disrupt the ecommerce and “product discovery” space by creating a mobile experience that captured consumption data. I figured that if we could be present at the “moment of consumption”, if we knew what product you were using, when, where, how often you had used it before, what you paid for it and a qualitative and quantitative indication of sentiment — we’d be sitting pretty. I would have data that would allow me to understand customers better — and therefore get them to buy more. I could start to improve the buying process. I could start to improve those horribly unpersonalised daily emails that local and international ecommerce is so reliant on.
With wine as a test bed, I thought we had an interesting insight into the reality of the market. 85% of wine in SA is purchased through a retail/supermarket channel. If you build a consumer facing product, you’re likely to fight for the 15%. If you build a consumer facing product that is actually a B2B platform (that you sell to retailers) — you can go after the 85%.
So the first business model was to target retailers: licence the technology into Pick ‘n Pay, Woolworths and others. Even if we had to exclusively give it to one. Plug it into their rewards programmes (where instant scale lies — most of these retail rewards programmes have more than one million active customers) and slowly expand into interesting verticals. It was something we got right with MyBEER and SAB but unfortunately that didn’t scale quickly enough on tight marketing budgets — so it only lasted nine months.
With the retailers, there could have only been two reasons we didn’t get anywhere. Either I was a crap salesman (I’d like to argue that’s not the case!) or we were just too early. I think the reality of retailers and ecommerce is that they just weren’t ready. Woolworths released their redesigned (pretty good looking) ecommerce platform in 2014. Makro also cracked this in 2014. Pick ‘n Pay, I think, is in the midst of an ecommerce rebuild. We were talking to them in 2012. And even though apps and mobile experiences were on the tips of everyone’s tongues — I fear we may have just been ahead of the curve.
The unfortunate part of the B2B business model is that is took us too long to fail — too long to get a final and definitive “no” from the retail space. If you’re building for Enterprise, remember that a two-year sale cycle is not uncommon.
“Timing” is a theme you’ll see appear over and over in the war stories told by entrepreneurs. It’s a critical factor, one that’s also often out of your control. Being ahead of the curve isn’t as sexy as it sounds. A startup of any nature needs to be on the curve — otherwise you’ll spend too much time convincing people they actually need you.
BUSINESS MODEL TWO. Sell the data. Even a year in, I was sitting on some pretty amazing data: consumption and sentiment data for wine and beer across a segment of high LSM smartphone users.
But there’s a big lesson to be learnt with data. Data is only as valuable in terms of what a company can easily execute off it. And consumption data is extremely hard to execute.
Let me give you an example. Let’s say we have a sample of 50 customers that buy Wine A at Retailer A. 45 of those 50 rated it YUCK. Yet the stock is still moving on Retailer A’s systems and they’re making margin. Are they going to change their buying patterns? I doubt it. Not yet, at least. Does that mean your consumption sample is not statistically significant? Or does it mean that “love” for a product doesn’t always trump price? I fear it may be the latter.
I saw research companies and internal customer intelligence departments get seriously excited about the data set we were collecting. At one stage we were looking at simple word clouds of phrases customers had used when reviewing Client Product A vs Competitor Product B – and building “taste” profiles that could theoretically drive marketing messaging. But I never saw anyone actually able to execute off it.
Data certainly is the new oil. Just make sure you can show people how to turn the oil into petrol.
BUSINESS MODEL THREE. Advertising. I’ll cut to the chase on this one. Advertising requires immense scale. As an African business model in anything except mass product markets, you’ve got your work cut out for you.
There’s a reason content portals and news sites are pushing millions of page impressions — that’s the kind of scale you need to make advertising work.
We built some pretty nifty ad tools, with the ability to target specific messages to people at uncannily specific times. You’re on your way home, you hit the SPECIALS button, it pulls back a list of wines on special at a supermarket on your way home. You’re drinking a wine, the app delivers you content specific to that wine (food pairings, tasting notes, a discount voucher for you next purchase or even a cross sell into a deli product that goes particularly well with that wine).
Here’s where scale plays its role. With 5000-odd users, we just couldn’t generate enough ad impressions to make the tools exciting. With the ad format where we delivered specific content to users during that beautiful “moment of consumption” — how many people were going to be drinking the same wine at the same time inside our community? We could deliver ten perfectly targeted impressions.
But only ten.
Add to this that media buyers just didn’t understand what we were selling. I don’t want to sound derogatory, but the media industry still works on things they can easily understand and bill: banner ads and search ads. If your “ad unit” doesn’t fit the mold — you’re going to have to do an education job as well as a sales job. Sales jobs are hard enough.
There’s a reason investors don’t jump up and down for advertising-based business models. They’re hard. They usually require lots of money (or time) to make money.
BUSINESS MODEL FOUR. Affiliate sales. Again, this business model provided a nasty little lesson in scale. We ran some affiliate sales pilots with a particular online wine retailer. Results were interesting, even quite exciting. Six emails into an untrained base moved around R100k worth of wine.
That’s a promising equation — if you own the ecommerce infrastructure. If you’re on a 7.5% affiliate fee (and that’s high for an affiliate) — do the maths. You’ve got to move a metric f$%! ton of product before anyone makes any money.
Affiliate models need immense scale to work.
And there we have it. Due to failing slowly, a fair bit of unfortunate timing and sometimes getting distracted by the sexy stuff… by the time we realised we had to build direct ecommerce (which I had avoided from the beginning because of its commodotised nature and the expense of the infrastructure) — it was too late.
Timing is everything.
Here we are back at “Timing”. During the Real Time Wine journey, my investors and I presided over three failed acquisition/merger opportunities. I say acquisition/merger because in reality, no one buys a one-man team unless the product has really shot it out of the park in terms of revenue and traction. But there are plenty of opportunities to fold that product into another stable, use it as a framework for a round of funding and crack on together with a bigger proposition.
One attempt failed due to some flip-flopping on loyalties and questionable handling of business ethics (although you have to keep reminding yourself that you probably didn’t want to partner with anyone that had these ethics in the first place). One failed due to culture fit and timing. The final one, a promising merger even if a bit of a long shot, also failed due to timing.
Timing is never anyone’s fault. It’s just a reality. And you’ll see it in the startup space from two angles:
Firstly, timing for finding a co-founder or business partner. In my first startup at university I had people lining up to get involved. We were students. Our risk profiles were the same. We didn’t need a lot of money, we wanted to have fun and we had tons of energy.
Since then, finding a co-founder has been an eternal struggle for me. I’m not sure if it’s the slightly (and I mean ever so slightly) elevated public profile — a friend calls this the “lone wolf” theory — people think you’re on a path that would never possibly include them. That you’re uber successful — they believe my PR! But it’s a more likely risk profile. Finding someone in their mid 20’s to mid 30’s who can make enough money on the side to pay the bond and feed the kids — and still put 90% of their time into a startup is tricky — timing.
You’ll also encounter timing on mergers, deals and sales. If the timing isn’t right between two businesses, that may have extremely different needs, wants, desires and shareholder forces — it doesn’t matter how sexy the deal may be, it’s unlikely to happen.
I’ve come to learn that deals require both momentum and hunger from both sides. Sure, every now and again you’ll find an exception that proves the rule. But I’ve rarely seen something happen where one side is chasing harder than the other. Unless it’s our good friend Timing at play again. I’ve definitely done some sales and deals in my life (like I’m sure you all have) where the initial chat and the actual sale are years apart. This is the tenacity “they” talk about an entrepreneur needing.
There’s not much you can do about timing — it’s the large “luck” portion of starting a business. Just keep having conversations — as many of them as possible. Be brave. I’m certainly going to be using the phrase “we should look at working together” a lot more in the next phase of my career.
Point your Startup internationally. Even if you can’t.
I remember this moment quite clearly. The original spec for Real Time Wine was designed for any wine drinker across the world. The database was architected by Prezence to allow for a very dynamic set of Country / City / Product Category / Product / Review / Rating data points. Then we hit currency and price as data points — how much a user paid for a product and where they paid it (restaurant, retail, wholesale). Currency and price are collective bitches.
Ignoring the retail vs. restaurant challenge, the first question was do we show the price in Rands (because we knew initial usage and revenue traction would come locally)? Or Dollars (to be more “international”)? If it’s Dollars, it’s pretty meaningless to local users — and we certainly didn’t have the ad budgets we thought we would need to compete internationally.
What if we allowed the user to put in the currency? But then we’d have to tie that user and product to a country and put a filter on the frontend to make sure we only show products or users from the country you’re in and the currency you’ve indicated. Throw into that melting point the fact that wine (as a product type) gets exported everywhere and you had a doozy of an architectural problem.
The edge case we explored was a South African travelling to the UK (we have some worldly SuperFANS so this is highly likely) and reviewing a South African wine available in South Africa but bought in the UK. Just how would we capture and present that data without making a mess of an already fragmented product set (there are roughly 15 000 wines per year on the shelves of South African wine selling stores).
In the “lean” mode we were in, currency got pushed to Phase Two and eventually out of Phase 2 and into the bucket of “never got done”. That was probably my fault. The solution is most likely a simple one, I just couldn’t see it. Perhaps this is symptomatic of our fledgling ecosystem just not thinking big enough…
From that point on, the reason I never focused on international audiences (aside from one failed attempt to target expats living overseas) was primarily because of the quality of our “Third World app” vs. the quality being generated in the “First World”, a horrible naming convention I know. I was pretty convinced that the app wrapper approach was just not good enough for international audiences.
I’ve spoken previously about the frustration some users experienced because of the mobisite within an app wrapper approach — that extra couple of milliseconds it takes to move between pages. Combine that with sketchy bandwidth conditions in South Africa — and Real Time Wine wasn’t the greatest experience unless you were on solid 3G/LTE or Wireless. This got us a lot of negative app reviews — for a product that was largely hailed as being fun, useful and a great example of South African innovation. That’s where my concern of competing in US/Europe was born. It was the worst rock and hard place challenge — I didn’t have the money nor the technical team to build native — so I got a bit stuck. Read back to the lesson on “channel focus”. If I had to do this all again, I’d go iPhone only.
Combine that experience with the complexity of currency (some smart ass is going to suggest an easy solution to me after reading this – I know!) and I basically never had the confidence to point this internationally.
The Internet works on sub-cultures and sub-currents. And successful (mega-successful) consumer orientated product startups by luck or design, manage to hit one of these. You can’t hope to catch a current if you aren’t out there.
Unless you have a very South African specific idea — point internationally. Be available internationally. Be usable internationally — even if you don’t focus on it in the beginning. If this scares you, try your hand at a B2B product first, we’re still learning how to think big in this country. Or at least I am.
Don’t worry about being copied.
I hear this a lot when I talk to entrepreneurs. Sign this NDA. I can’t tell you what we’re up to — it’s too good. I don’t want to release the product till this is perfect, otherwise someone is going to copy us.
All rubbish. And usually, in my opinion, this language is a terrible sign to potential investors.
US and European startups plan to get copied. Their original strategy docs usually have reaction plans for when it occurs. They take it as a sign of market validation (sure, they’re pissed off, but we live in a Xerox world — with great freedom comes great ability to rip something off). There are even businesses like Rocket Internet that have made hundreds of millions of dollars simply by copying other products. If you aren’t copied by China — you’re probably doing it wrong.
We were copied within about six to ten months of launch. In some ways, we were copied pixel-for-pixel. I got pissed off. I know Prezence was livid (agencies still come from the culture of creative — where stealing ideas is criminal, not the culture of product — where borrowing ideas, even in little bits, is how it’s done).
I took a screenshot. Patted myself on the back. High-fived my imaginary co-founder and moved on.
Don’t worry about being copied. Unless you’re sitting on a cure for cancer, success lies in getting out there and iterating quickly — not in hiding everything until some big reveal (a reveal you usually won’t have the funds to underpin).
Be tidy. Just in case that deal comes through…
Over the course of the startup, I spent close on R150 000 on a combination of lawyers, accountants and :managing the books: (a hefty slice of total funding raised). I remember having a big fight with AngelHub about this, because both the lawyers and accountants (and the quality of them) was their idea. “Do this properly,” they said, “keep all your financial stuff clean as a whistle, it’ll keep your investors happy.”
Looking back now I can tell you two things with fair certainty. I still think I spent too much — but that’s because I failed. If I had succeeded that number would have faded into obscurity and I would have thanked my lucky socks I had done it right.
During one of the merger deal discussions, we started to open up company books to each other — and the other company was in a complete mess. It was kind of nice to be the small upstart that had a professional set of books that didn’t need any rush work.
Deals are hard enough. Why would you want to add complexity and barriers-to-success by having a messy set of accounts and a shoebox-full of till slips? You don’t want to do that. A good deal with good people comes along, you do everything in your power to make that as smooth as possible — because it’s very rare that one company truly needs another.
So get those management reports out monthly. Spend a little on an accountant and lawyer if you can. You don’t have to bust the bank — but go look at our idolized US/Europe counterparts, sure they’re raising more money than we are, but they’re doing things properly. I’m sure your dad’s a great guy — but there’s a stage fairly early on in a startup’s life cycle where he’s probably not the right guy to be running your books.
Conclusion: Startups. The messy and joyful art of getting your Street MBA.
The “Street MBA” — borrowed this phrase from a friend, Andy Volk, who I’m sure borrowed it from someone else. It’s such a good phrase — one that I hope replaces “serial entrepreneur” and “entrepreneur enthusiast” — some great new vomit words the Internet has brought into our lives.
The “Street MBA” celebrates what you learned, not what you got wrong — because getting something wrong is pretty much the only way you learn.
Failure isn’t a bad thing. South Africa has a horrible stigma around failure. Where US/European startup communities celebrate it and wear it like a badge of honour — we hide under a rock and pretend it doesn’t happen. I’m not going to lie — I’ve been pretty worried to write and release this, because of that stigma. Hopefully this content does something to move us towards a more accepting, better culture around startup failures. Failure is a couple of extra course credits in the Street MBA. In the US, most investors won’t even look at you until you have a couple of failures under your belt.
The Street MBA is a lesson in business, a lesson in stress, customers, friendships, partnerships, revenue and product. A Street MBA teaches you how to recover from being punched in the gut daily. As I write this, I’m smiling. Which is a far cry from the miserable sod I’ve been during the slow death of Real Time Wine. That’s because it’s ok. I’d do it again. In a while.
I hope you enjoyed this. It sure provided the catharsis I needed — so the rest is up to you. Get out there and start something.
Image: StockMonkeys.com via Flickr.