What must a startup do to survive?

In the past, the blog material we talked about the project SHOPPILOT ( History of the project SHOPPILOT: From a pack to a startup ). We met with them being in the Accelerator of IIDF. This time we want to share the knowledge that the learning process in accelerators gives. We will tell you how in 3 days we were given the basics of start-ups on the Web Ready GenerationS 2014 program, which was held on the basis of IIDF . From this article, you can find out what stages a startup goes through before becoming a business, how best to prescribe the business model of a project to attract investment and a few life hacks that will help a startup survive. First, let's get a little “bored” over theory, and then “cheer up” real cases.

On the first day of training, Dmitry Kalaev, director of acceleration and educational programs at the IIDF, told how a startup should calculate the amount of necessary investments for an investor, what to do if the project has a complex business model, and also focused on other things that are useful for a novice entrepreneur, supporting the theoretical part practical case studies of Accelerator FRIA.

One of the basic tenets is: a startup is not a business. You can call it business in the bud, but there is a very high probability that nothing good will emerge from it. In order to understand whether a profitable business can turn out from an idea at all, a startup needs to be driven out in several stages.

In the presentation of Dmitry Kalaev there is such a picture:
')


The first two stages of a startup using this methodology are hypothesis and testing . And if in the process of testing the initial idea is not confirmed, then the startup has to do a pivot - reversal, build a new hypothesis and test again, and so on until it is successful. The construction of a hypothesis, in essence, is a search for a business model, during which the entrepreneur must confirm that there is such a problem, that his project helps to solve it, and that there is a solvent audience that really needs it. The main evidence is the purchase and / or installation of the product by customers.

In the process of testing , the viability of the project's economy is checked, here we must confirm that a business can be built on the product. The result of testing should be such a unit-economy of the product, in which we spend on attracting a client more than 3 times less than we earn on it (for more details, read the article “ Growth Hacking: How to earn more, using metrics correctly ”) . Also, the economy needs to be tested for ability to scale.

If everything is successful, the process of pumping traffic into the project begins to increase the client base. At this stage, seed investments can already be attracted. Between the stages of testing and pumping - a crucial moment for a startup. The first two stages differ from the third in several ways. First, first, the team often does not have clearly defined roles, all team members can perform any functions. Only at the third stage do team members usually have clearly defined areas of responsibility. Secondly, mistakes in the first two stages are even good, they allow a better understanding of the product, business, and concept. Plus, the losses from them are quite small. In the process of pumping, the packs are already much more expensive. And thirdly, when the economy is confirmed, the business model is clearly stated, the project is already starting to realize its main mission - to make money.

In IIDF, the business model of each startup is described by several components based on the “ IIDF Lean Canvas Model ” methodology, created thanks to the IIDF ’s own practice and the A. Osterwalder, E. Maurye and RIS Ventures models well known in the market. Components of this model FRIA:
• problem
• technology
• product,
• market
• channels
• "sales funnel",
• customers,
• product economics.

The simplest business models know everything: b2b, b2c, b2g, more complex b2b2c.

But many companies are forced to operate in several markets and target segments at once, which makes their business model two or more times more complicated. An example from the past set of the accelerator: E-transport company (a mobile application that shows in real time when a particular bus or tram arrives at a given stop). Their customers are both individuals who install the application for free, and legal entities advertisers who sell this audience. They have different channels, different economies. In cooperation with individuals, it may be planned as unprofitable, but at the expense of advertisers the company covers these costs. According to this model, Twitter, Facebook, and so on live: they give us their services for free, but as an audience they sell us to advertisers. Such companies have to work on two fronts, as a result, ETransport has two business models: individuals with their own channels and the cost of attracting a client, advertisers with other channels and another economy.

Another case of a complex business model that Dmitry spoke about: this is the company SimpleOrder - a system for automating the activities of hostels and small hotels (from 10 to 50 rooms), in the basic version - free. But when the owner of the hotel uses it for about a month, the company offers a paid service of automatic updating of the available room stock on Ostrovok.ru, Booking.com and other booking sites. And it works, because there are frequent situations of re-booking the same number on different sites. A man will come, he will have to be settled, an awkward situation may emerge, and the main thing that has made such an oversight hotel in agreement with booking.com-type sites is responsible for the client’s accommodation, which can lead to direct and substantial losses. The top component of the business model is the lidogeneration: hotels easily agree to use the free service, and the next step is to create a value proposition for these people, thanks to which some of them are converted into buyers.

Dmitry Kalaev's speech on Web Ready:



What is worth knowing to those startups who have such a model?

You will not be easy. When there is just a product that you directly sell and collect money, then the hypothesis of this cycle can be checked very quickly. When this is a complex model, you first need to generate leads, then pass them on to someone else, and someone else will pay. Until the whole chain is formed, it will take quite a long time, and in the end the result can be unsuccessful. We'll have to go all over again. Dmitry's life hacking: you just need to understand which of these models to test in the first place, focus on it, and emulate the rest of the business model manually (for example, instead of an automated contact transfer interface to the auto dealer, make the call a real person).

An example from the first set of the IIDF accelerator is the Online Dealer project (a service for booking new cars from official car dealers). The hypothesis was this: car dealers, most likely, are ready to buy leads - to pay for attracting a person for a test drive. The service was a kind of widget with specific models of cars, they wanted to hang it on portals like Avito.ru , Auto.ru. To test this obvious hypothesis, it was necessary to negotiate with dealers and portals. By the time it took 2.5 months, that is, two weeks before the end of the acceleration cycle, they had just started testing. They received the first money from the client only when the acceleration was over. What conclusions can be made? Understand that dealers are willing to pay for attracting leads for a test drive, it was possible with the help of Customer Development - just by talking to them. When testing a hypothesis, it is necessary to correctly determine exactly the place that really needs to be checked in order not to lose a lot of time for nothing.

Consider such components of the business model as channels and market .

Channels are not only advertising locations, but also methods of communication with customers, their attraction, as well as funds aimed at it. It can be both personal sales and targeted advertising. Prescribing channels to attract customers is necessary, first of all, for the investor: he wants to know what gives money.

For example, there is a Yandex.Direct channel, we prescribe keywords for it, channel capacity, conversion, transfer price, price for attracting one customer, how much we earn on this channel and the amount of investments that need to be attracted to cover the channel. Then you can discuss it for a long time, look for ways to optimize. But in any case, if you explain where and why the money will go, it will cause investor confidence. To know the channel capacity and conversion, you need to work with funnels: if the channel capacity is 500,000 people, and there are only 5 buyers, something obviously happens wrong between these two numbers, this needs to be clarified. Maybe Yandex.Direct doesn’t have that message, or the benefits of the product aren’t disclosed on the landing page.

A few key points when drawing up a funnel:

1. It is not necessary to collect all the channels of attraction into one funnel, otherwise you get an unreal picture that shows nothing: the average temperature in the hospital.
2. It is necessary to go all the way to the site visitor: problem points can be hidden in the interface itself (registration does not work, disorienting the menu, etc.).
3. If so far not all the figures are known, you can put the average market ones to calculate the project's economy at least approximately.

For the investor, the correct calculation of the market is no less important than the calculation of the promotion channels and the funds necessary for them. There are three market sizes:

TAM (Total Available Market) - the total volume of the market for similar goods;

SAM (Served Available Market) - the available market volume, that is, the potential market, taking into account competitors who already work on it;

SOM (Serviceable Obtainable Market) - attainable market volume, 100% of your turnover.

An example of incorrect market calculations. A few years ago, Dmitry, using his own money as a partner of the RedButton Foundation, considered an investment in a rental service that he wanted to replicate. To the reminders that this market is busy with AirBnb , the guys responded that six months ago they had already made such a service and sold it, now they want to do it again. In an attempt to figure out how they did it, they were asked to count the markets. They counted that the market for daily rent of apartments and cottages in Russia was $ 1.2 billion. They assumed that 15% of the market is booked via the Internet, SAM turned out to be about $ 200 million. And if you take a million-plus city other than Moscow and St. Petersburg, amount to no more than 10% of the market, and the maximum SOM is about $ 20 million. It would seem that normal money. But they forgot one nuance: the service earns a commission from 5 to 15%. It turns out that TAM - $ 129 million, potentially to the whole of Russia - $ 19 million, and in one city - a maximum of $ 2 million, for which you no longer want to fight. Moreover, in the process of field research, it turned out that in this particular city less than 10% of real estate is booked via the Internet. The saddest thing is that most services with a similar situation continue to count in the same way.



An interesting example is the dual market of the graduate Accelerator FRIA. EasyTen (a service that offers to learn 10 foreign words every day) has opened for itself two audiences: application users and companies that can sell these same users as customers - educational agencies, foreign language schools, tutoring centers, etc. Accordingly, the market that we end up with two pieces: first we sell the solution to people who want to learn languages, increase the base, and then sell the interested people to the agencies, in general we can get $ 1.7 billion.

Another important question for an investor is: how will they be better than competitors? EasyTen rebuilt from competitors when dealing with investors as follows: in order to bring LinguaLeo to Spanish, Chinese, Portuguese and other markets, you need to spend a lot of time and money on localization and adaptation. Since in EasyTen, users learn only words to enter a new market, in fact, only a month is needed to introduce new pairs of words.

There is another example of GeneGuard - a script that is embedded in a web service, and diagnoses hacker attacks and makes maximum efforts to prevent them. The guys considered their market quite simple: there are only 200 million people in the world who have such sites and who may need this, and 6.7 million of them are willing to spend money on security. Every year, about 675 thousand companies already face attacks. They differentiated from competitors at the lowest price of this solution due to the maximum specialization in protecting the site from hackers. The rest of the services in this area do besides this monitoring of the site, and many other services, so the price of their service is much more expensive.

New stories about the acceleration process and start-up projects that we met here - read in future publications here and in our blog on the site .