My Twitter friend @sebzachrau asked me recently how does one build an Uber (or Twitter) type of app from scratch. If you are a pro in the field, you will probably not be interested in what will follow, all other folks having "an idea for an app" may though, so here we go 1/n

Sebastian asked me as well if there is one Twitter Guru for these things (building apps) as say @GuruAnaerobic is for health advice. I honestly do not know, probably. I asked one question on @Quora on early stages & I got a decent answer, but I did not try the same on Twitter 2/n

Note I had worked on some mobile apps projects for @Airbus (on workshop tablets and on UI5 technology), but I had never built an Android or iOS app. So I had to learn many things to do these, hope these learning can be useful for some of you reading this 3/n

Also, before I start going into the details I would like to tell you that the market for mobile apps is pretty similar to the market of books, really. Both in the turnover of these market and in the numbers of apps/books published. I checked this recently, but the bottom line 4/n

... is that your odds of making it big with an app are pretty similar to your odds of writing a bestseller book: pretty slim as you may have guessed, nevertheless success is not as unlikely and random as wining a lottery because it (partly) depends on your performance, partly 5/n

... on sheer luck. So in our case, my sister, who has plenty of ideas, had one idea for an app: a marketplace for neighbors sharing food they would have cooked. I really liked her idea, and she was willing to help put it in place with funding, marketing support and more ideas 6/n

I had a co-founder, but not a technical one. And me, I'm only slightly more techie than she is, so we would need some help. So I decided to invite my techie friends over dinner and pitch them the idea. 2 out of 4 of the folks I invited showed up, and liked the idea as well... 7/n

One of them is still my partner, the other would jump out the boat later on. Anyway, the three of us did a product backlog and "sized" what an MVP would cost us to build, with all the features we wanted, it was slightly more than 300 development days. As per the law of the... 8/n

large numbers, that turned out to be a pretty accurate estimation: what you would overestimate on one task, you tend to underestimate in some others. We also agreed to grossly price that development work with the average of an Indian and a French developer salary shall we... 9/n

later decide to create a company, determine how much equity should correspond to each one of us depending on the tasks (user stories) one had developed. We also agree that the value would be degressive depending on the completion dates and started working on this project 10/n

& note also we priced the project management tasks, idea, branding, legal admin fees etc. at roughly a similar price to the one we had given to the development work. Eventually, when we would decide to incorporate (later on) a French "commissaire aux apports" ~ (chartered... 11/n

accountant) validated the valuation, so I guess we were pretty accurate with the same. We did not seek the maximum possible valuation on the other hand, as it would hamper our chances of onboarding a purely technical, experienced co-founder later on. At this point all of us 12/n

... were on payroll, and working full time jobs as we did this as a "hobby". And we had to learn on top of it how to do this, as it was sort of the first time we tried to do something like this. So there lots of late sessions around pizzas and some reading to do to catch up. 13/n

Note also that we had ruled out to do a hybrid app. We were kind of optimistic and ambitious, and we considered a key success factor to have things like a payment gateway, which would mean an app slightly more complex than what you can code with Apache Cordova @apachecordova 14/n

We also considered mobility paramount so we ruled out doing a marketplace with something like @sharetribe. In retrospect, that may have probably been a decent idea to at least "test the waters" before deep diving. We did not want to fail because we would be in "Mediocristan" 15/n

so we want all the hard way. My techie friend had some decent Java experience and we downloaded Android Studio, where he would so some progress, but not enough to meet our targets, we were lagging behind, but getting thrilled by the taste of building something we liked... 16/n

I watched some tutorials on YouTube and build the first version of the app in WordPress, that we would use to pitch around the idea. I also build a first SQL DB, but we would later scrap most of it as we would appeal to folks with more experience on these fields to help us 17/n

Note also we just wanted a working app, not something particularly fancy. We gave a priority to the Android app as we guessed (so far, correctly) that it would be more popular than the iOS one. For the backend, we chose to use PHP with Rest Server. And we posted an ad for... 18/n

getting support building the APIs/webservices. We got a decent amount of replies, and actually one guy in Northern India (whose name I will spare here) and who apparently knew well the topic made us the kind of offer we couldn't refuse. I was quite confident in our estimates 19/n

... and to some extent, pretty sure he was "low-balling" us. But on the other hand, we said to ourselves that if that was the case, we did not have much to lose. So we worked with these guy for some months, and he did some actual progress, especially in the most visible part 20/n

... that is, the frontend of the Android app even if we had originally launched our "RFP" to outsource the backend development. Outsourcing most of the development may seem not a recipe for success, but the number of startups that have done this on early stages is staggering 21/n

Anyway, we would indeed realize soon that the guy from Northern India was indeed low-balling, as he kept asking for more money e.g. to use the repo we had created in https://bitbucket.org/  to store the code. So somehow ironically, we were being "swindled" but we saw as well 22/n

... that this way of working could work out. So we started looking for a trusted folk in India to work in our iOS front-end development. On top of it we started the process of incorporating the company, my sister and I would down pay some 80% of the money we raised among FFF 23/n

... 20% or so was provided by the other "partners in crime" and some friends. The amount we gathered you can guess it if you read carefully so far. More or less the cost of the average car in France... And I was rather driving this company than a new car! We incorporated... 24/n

As a French SAS, that is the most "scalable" type of company in France. I was in Toulouse all this time, in case you were wondering. Also, more or less by this time we would find @amitsingh, founder of HummingWave and the trusted folk I mentioned in 23. Amit was a former... 25/n

Infosys employee, same as I (I requested a sort of "sabbatical" when we incorporated the company) and we had common acquaintances. Amit is as honest and competent as they make it and I only wish we had started working with him earlier. So he helped us building the iOS app... 26/n

and eventually, also, he would also complete the Android app pending job where the others had left it. Fabien also joined us as CTO thanks to @d2villers, which was a blessing. He was off from having created his own startup and accepted to work mostly for equity (no way we... 27/n

could have afforded him otherwise! Fabien rebuilt our APIs and our webpage with the pretty result that you can currently see: https://www.okakiben.com/  Fabien is the best full stack developer you can find! So we were lucky indeed. After we had these cracks with us building the 28/n

... apps & getting them published in the App Store and in Google Play was a piece of cake. So this was the story of how we build the apps, I have blogged about it here if you want some more details. Feel also free to ask me any questions if you want. 29/29 https://www.huffingtonpost.es/ignacio-oliveras/diario-de-una-startup-el-mejor-momento-para-emprender-7_a_23339997/