Pivot: pitfalls and conclusions on the example of the startup MoneyHero

As practice shows, the business idea in most cases comes to its real incarnation at least after significant reworkings, and often - only after the pivot - a full restart of the startup, accompanied by a radical change in the business model. This is due, primarily, to the fact that at the stage of forming a business model, the founder does not yet know exactly how it will be implemented in a particular business segment. In other words, if you have not worked in this area before, are not “immersed” in the subtleties of doing business in a particular area, this very area will surely (and quite seriously) adjust your business model or even put before you the impossibility of doing business in this form, in which he originally thought.

One of the illustrative examples of successful pivot: Groupon, which was originally conceived as an alternative to Kickstarter for non-commercial projects (in this form, investors did not like it, who were not sure of the possibility of monetization of the business). As another illustrative example, the story of Twitter is often cited: before coming to the idea of ​​microblogging, its founders were developing a platform for podcasts Odeo, which in the end could not compete with iTunes.
')
The acceleration program of IIDF was also attended by companies that faced similar problems and with dignity emerged from the current situation, one of them is the startup MoneyHero. Using the example of its history, we will try to figure out how (using the example of other people's trial and error) to develop approaches that will help you evaluate your chosen business area for the viability of your idea, check the idea as soon as possible and not make common mistakes associated with startups at this stage.

What came to Accelerator

At the time of arrival in the Accelerator project MoneyHero was 6 months. “We went to the third set with the project - an aggregator of financial loans for individuals,” says the project’s co-founder Anton Utekhin. What was the essence of the idea? The scope of microcredit is such that the application is likely to not approve. To increase the chances, it is better to leave a request at once on several sites of different companies. On each of them you need to fill out the same form from scratch. “It’s not very convenient for many people to be lazy. Therefore, we decided to create a website on which you fill in one, maximally complete questionnaire, and already it is sent to all microfinance organizations (MFIs). We focused on companies that issue loans online. ”

As conceived by the founders, microfinance organizations had to pay the service for attracting customers from 1 to 3% of the amount of the loan issued. The risks of non-repayment of loans were on the side of the partner, and in this respect the model was interesting.

Before joining FRII, the site of the service looked like this:



To test the potential market using the Lean Startup methodology, the team needed to calculate CPA (Cost Per Action, cost of target action) and ARPU (Average Revenue Per User, average revenue per user). In the case of MoneyHero, these metrics depended on% conversion to registration,% conversion to completed questionnaires, and% conversion to loans issued. The traffic was provided by contextual advertising. The team calculated that with a CPA of 30 rubles, it was possible to attract an audience of 1,000 people to the site with 31% conversion to subscriptions and 7% to completed questionnaires. “These were very good numbers that inspired us.”

Testing the first business model took two months

IIDF immediately set the bar high - by the end of acceleration, it was necessary to earn at least 1 million rubles a month to talk about a seed-round of investments. To check whether the project will be able to meet this requirement and whether this product is necessary for customers at all, MoneyHero posted on their website a questionnaire that they created using a third-party service. All the applications received by the team were going to be manually entered into the questionnaires on the partner sites.

Thus, the company wanted to find out the following things:

1. How many completed questionnaires are converted into loans;
2. How many users who filled out the questionnaire will already be registered with the MFI.

The last point was especially important because during the negotiations with the IFIs a limitation arose - all but one of the companies were not willing to pay for repeated appeals from clients. This questioned the economics of the project.

In accordance with the new logic, the team posted a questionnaire on the site and updated the site itself in order to increase the conversion to the questionnaire. As a result, he acquired the following form:



The test results were initially not so bad, considering that the questionnaire was quick to do:



“And we began to transfer these 52 questionnaires manually to the websites of partners. It took our employee 3 full working days. In the process, we were faced with the fact that almost all MFIs had SMS verification of telephone numbers. Therefore, we had to call each client and ask which SMS code he came to. And some companies had 2 sms-codes ... ”, says Anton.

The company found that in 60% of cases, clients were already registered in large online MFIs.

Later, the Spartup managed to negotiate with one of the MFI partners about API integration, which took about a month. "When it ended, we tested the audience automatically and found out that 92% of our site visitors are already registered on this partner's site." Of those who were not registered, the loan was not issued to anyone.

“The main mistake we made at this step was that the MFI believed that they were issuing loans to an average of one fifth of all incoming clients. Later, these figures were not confirmed. And initially our economy was built on them and on the assumption that we will also pay for repeated applications from the same clients of MFOs, ”Anton Utekhin.

As a result, the company came to the following facts:

1. Of the eight partners, only one agreed to pay for repeated applications via the MoneyHero website from the client that already exists in their database.
2. The contextual advertising channel was practically “chosen” by large online MFIs. The audience from alternative channels (social networks) did not leave the questionnaire.
3. The total income from those customers who were approved to pay the loan did not cover the total costs of not attracting the entire client base as a whole.
4. Economy “Application forms for a loan” was interesting only under the condition that the volume of traffic is at least 25 thousand per month, and the conversion into loans is 7%. “Then we could achieve a turnover of 1 million rubles a month - the minimum amount that might be of interest to investors and IIDF. The issue of traffic could somehow be solved, but the conversion into loans on a general scale was almost 0%, ”the project co-founder summarizes.

Findings:

If you are not completely sure that you know the “pitfalls” of the sphere in which your project will work, try (and, preferably, the sooner the better) to test the project in practice: your quite optimistic financial forecasts may ignore a number conditions that initially (in the absence of special knowledge of the industry) it will be extremely difficult (if at all possible) to foresee. The MoneyHero example visually demonstrates how logical at first glance the circuit in reality turns out to be unviable.

Nevertheless, the MoneyHero team, familiar with the Lean Startup methodology , did it right, trying to create a minimally viable product (in the terminology of Lean Stratup - Minimum Viable Product, MVP or, in a more modern interpretation, minimally viable transaction, Minimum Viable Transaction, MVT ), which allowed them to save time and resources for further work in Accelerator.

“We realized that further work in this direction now is just a scaling of losses, and decided to concentrate on another market of this business model — loans for business,” Anton Utekhin.

Pivot first - Testing the new market took two weeks

The following facts spoke about the prospects of the loan market for legal entities:

1. The volume of the microloan market for legal entities according to the data at the end of the first half of 2014 amounted to 18.2 billion rubles, which is less than loans for individuals (28 billion), but also not bad.
2. The commission was the same - 2-3% of the amount of the loan issued. But! The average check of loans for individuals - 10 thousand rubles for one or two weeks, for legal entities - about 300 thousand for a month or two.
3. There are no online loans in this market, but companies needed only contact of the borrower, they met with the client independently.

Using the lessons learned earlier, the team negotiated with 8 companies for several days and agreed with the three largest partners (able to close a large geography of loans) that will attract customers for them. Also, MoneyHero bought a new domain ( www.denginabiz.ru - is now disabled, since hosting is used for another project) and collected a new landing page:



After they started advertising, they collected 300 suitable applications for 120 million rubles and gradually transferred them to their partners (in the form of MS Excel tables). Following 2 weeks of testing, the following data were obtained:



The startup had a clear problem with issuing loans. It was still possible to work with conversion, but the team was limited in tools with loans. The company decided to analyze the situation:

1. From the first day, the team tracked the sales funnel and constantly asked the partners for the status of the submitted questionnaires. Status was tracked by weekly cohorts:
   
   
   
   Logically, over time, clients from the “In Work” status should have moved to the right, but this did not happen. Anton explains: “We clarified the information. The main reason - "customers do not get in touch." All the next day we called these customers. From the first call, 70% of clients answered us and answered that they did not communicate with the MFI. The partners did not deceive us, but the quality of the work of their operators was doubtful. ”
2. The startup attracted customers who filled out loan questionnaires for a total amount of 120 million rubles. The volume of the entire market is 18 billion rubles. “Roughly speaking, in three weeks we grabbed almost 0.7% of the market.” This fact, in conjunction with paragraph 1, could speak of a large disproportion: the demand for issuing loans greatly exceeds supply.
3. The main reasons for the non-issuance of the application were the refusal of the security service partners and not the passage of scoring. “There can be only one conclusion here - for these partners, our audience was not of very high quality. There was an option to go offline. We didn’t do that, since large investment costs were required, ”the co-founder of the project comments.
4. Another weakness of the business model was that payment was made again only for new customers who were given a loan. At the same time, out of 300 applications, 10% were repeated - better than in the case of individuals, but also a dead end model of development: at some point the channels for attracting customers would be exhausted.
5. In fact, the company did not earn.

Findings:

The transition from the B2C sector to the B2B sector is a rather popular pivot practice, not devoid of its own intricacies, but, nevertheless, on the whole is quite viable. First of all, this is usually due to the likelihood of reducing pressure from competitors (if we are talking about e-business) - mainly because of the satiety of the B2C market. On the other hand, as John Bischke, CEO of Entelo, rightly noted : “If you find and solve a problem that is relevant for thousands of B2C clients, you’re unlikely to succeed ... But if you solve a problem for thousands of B2B clients, you will have a very viable company. Yes, even if you solve a problem for a hundred companies, your business will be viable. ”

This is one of the main advantages of the B2B market: it is easier to segment, it is easier to reach (it took our heroes three weeks to cover 0.7% of the market), clients in this area operate with larger amounts, and their solvency is usually higher (in our example, the average loan check for legal entities turned out to be 30 times the average check in the B2C segment). Such a course change can be attributed to a change in the user segment (customer segment pivot) and business architecture (business architecture pivot) according to the classification of pivot from Eric Rees.

Nevertheless, ignorance of the same “pitfalls” of the microcredit sphere did not allow for the planned return from the idea (we must pay tribute to the MoneyHero team - this time it took them two and a half times less to test the new business model).

Pivot second - Testing the new business model took three days

“We began to think, what else can you try to earn? An interesting topic seemed to us to be lending to individuals. ” Testing the business model of the service NoCredits was carried out for three days. The first site was made with the help of the designer, the second one got traffic there. On the third day, the team processed these customers to figure out how many of them would take advantage of the offer, and calculated the funnel.
Landing in a hurry looked like this:


Full version - www.nokredit.ru

They were going to earn money on clients (500 rubles per decision) and lead generation for banks. The project was suspended because:

1. The sales process was too long. “Most people in our country do not yet understand what lending is. They need personal consultations, but even they do not always help to convey meaning, ”explains Anton Utekhin.
2. According to the financial model of the project, in order to earn at least 50 thousand rubles on it, it was necessary to attract 1,000 leads. And this is without CPA. It would be possible to increase the profitability due to lead generation, but ..
3. Only 1 out of 15 clients who have switched to the bank’s website have filled out an application for a loan, and have not approved it.

Findings:

The MoneyHero team found out in practice that it was impractical to develop the project in the initially chosen direction. This situation is extremely painful for the team - primarily psychologically: a complete change of course calls into question the meaning of a startup.

However, even in the face of such radical change, you are not alone. A curious example (as well as one of the options in case you feel that you have encountered something similar) - the project 30 Days of Honesty , created by the co-founders of the Dabble service, in which they met, communicated, worked and documented what was happening to ultimately decide whether to live the project or die. The story of Dabble is also attractive because it allows one to draw conclusions from a similar situation (and, if not to emerge victorious from it, then at least not to earn a serious nervous breakdown).

Pivot third: two weeks before the end of acceleration

“On August 29, we finally established ourselves in the thought that none of these projects could earn. We decided on the third and most ambitious pivot: our new project was completely unrelated to the topic of finance. ” The team came up with the EasyScript project - a service for rendering scripts for telemarketing and incoming calls.

“The message was this: what could be true of all projects sitting here? What we personally had a lot of difficulties? The answer was simple - phone calls. We ourselves used the mind map and logical conversation patterns. ” The conversation scripts created in this way were convenient and made it easy to navigate the conversation. “And we thought, why not create a cloud service with which you can prescribe phone sales scenarios? And then show them to the operator. " The task of the service is to process incoming and outgoing calls more competently. How to do it? First, using the service, a script is created in the form of a diagram with possible response replicas to the client. What problem does service solve besides added cost in the form of a simple interface? Usually we see only two numbers - the number of calls and the number of sales. And EasyScript allows you to open the veil and find out what is between them - why the calls are interrupted and where in the conversation.

For example, you can follow the conversation of the operator with the client in the service according to this scheme:



The team began to test the market:

1. First of all, the team printed its scripts and turned to other projects, presented them the idea. “Already at this stage we found two potential customers. We implemented the EasyScript work model on the TypeForm polling service. ”
2. Only after that the team took up the site. Landing was assembled on the Wix designer and as soon as it was ready, the project participants immediately told about it in their social networks. All registered customers are interviewed by the team to identify the need for telemarketing.
3. After creating the demo, customers selectively received free weekly access. The project team helped them develop scripts and transfer their scripts to the service.
4. A week later, customers were informed about the price of use (full price - 3,000 rubles, now - 1,000 rubles, since the project team has not fully implemented all the functionality). The goal is to confirm the value of the service. Free it is interesting to many, but because everyone needs paying customers. "Currently, out of 56 customers, 6 are already paying, and another 23 are testing - a good conversion."
5. “We also monitor customer engagement by examining the number of scripts in their accounts. The client has created a script - he is interested. No - we ask why and offer to help with the start of work. ”
6. “We tried to evaluate the market, but we had difficulty. How to understand how many scripting services are sold in the sales consulting market? We know that the consulting market volume is 34 billion rubles, we know that the number of relevant queries in Yandex is about 10 thousand, and what is between them is not completely clear to us. ”

In the near future EasyScript plans:

1. Refinement of the service to the level of the promised functionality, and checking the scaling of the project.
2. Formation of new sales channels - through coaches and consulting companies.
3. Integration with CRM (AMO, Megaplan, Speed ​​Dial). With the help of the service it will be possible to enrich CRM data.

“Even if our project once again does not“ fly ”, we will continue to use the scheme developed during the acceleration period in the IIDF, as it greatly reduces the time and cost of launching the project”:

1. Manual sales, check ideas;
2. Landing site designer;
3. Traffic;
4. Testing for the audience;
5. Product refinement;
6. Statement of TZ on a full-fledged site programmer.

And what is the result?

Using the example of acceleration (and three pivot's) of the MoneyHero / EasyScript startup, the following conclusions can be drawn:

1. Strive to develop a scheme that will test the business model as quickly as possible. “We tested the first business model for a month, the second for 2 weeks, and now we have three days to get the first results.”
2. First, try to sell the product, and then program. There may not be a complete product, but its value is better to confirm in advance, so as not to waste time and resources in vain. It's a shame to make a good product that no one needs.
3. It is important to immediately confirm that the product will be bought. Use Erik Rees's Lean Startup approaches or other methodologies that use the concept of a minimally viable product. “We decided to try to sell an unfinished product for a smaller amount. To confirm the demand, the site can be made on the designer. If the idea goes, then go ahead in the TZ and the development. ”
4. Only numbers can tell how effective the project and the idea are; subjectivism is better to put aside and count the metrics.
5. It’s very hard to kill non-performing projects, but sometimes it’s better to stop on time. When to do this - prompt the numbers. For example, EasyScript hypotheses sounded like this:
   
   • Will the first 5 customers appear?
   • Will the project be able to provide conversion to sales on scripts in 30% of the site traffic?
   • Will buy 3 thousand / script?
   • Where to get 700 people per month?
6. Ambitious goals are better than realistic ones: 500 thousand with a goal of 1 million rubles is more than 110 thousand with a goal of 100 thousand. In addition, large numbers immediately indicate bottlenecks in the financial model.
7. While working on a project in an area whose “pitfalls” are unknown to you, strive to minimize the number of external elements in your business model and the dependence of the results of your work on them. “After working on the project“ Money for Business ”, we decided that it is better to create a model in which key indicators depend only on us. In our case, we were quite dependent on partners. Therefore, we sell the new product ourselves. ”

Ps. Apply for full-time acceleration in the IIDF here . For those projects that are not yet ready for MVP, or have difficulty in formulating an application, there is a free online program of preparation for the presentation to investors - Preaccelerator .

The story about the project MoneyHero was prepared jointly with Maria Lyubimtseva (content specialist, IIDF).