The Kamino Lead Generation platform employed 186 students and part-time employees in first 6 months and has been built using the HOOKED model and gamification principles to ensure retention. We did 15 user tests. The platform reduced the cost per lead (CPL) for client from $2.18 to $0.34.
SlidesLive is a company focused on recording academic conferences. My task was to try to find a solution that would make the work of sales more efficient. The CPL (cost per lead) at the beginning of this project was $2.18. The goal of the project was to design a platform on which Czech temps could work and generate more leads at a lower cost than a sales team in the US.
This project was initiated by me after a discussion on business processes with the CEO of SlidesLive. I was responsible for the whole project, from the idea to the working prototype.
I spent 2 weeks gathering data and conducting in-depth interviews with the sales team to find out how it currently works. I mapped out the current situation. We knew we wanted to create a tool for Czech students, so I conducted a survey among students.
Key findings:
I've found that the SlidesLive Sales team spends about 50% of their time looking for conference sites to email to. When I calculated the CPL (cost per lead) based on their salaries, I found that the CPL was $2.16. Based on the average hourly rate in the country, we should be able to hire temps (mostly students) to look for leads for about $0.3 - $0.5 per lead. That could mean savings in the millions per year for SlidesLive.
Challenge
Design a product that would allow Czech students (and part-time employees) to find potential customers for SlidesLive.
Challenge
How to motivate people to work and enter the right data?
Answer
Hooked model by Nir Eyal
Nir Eyal investigated how the habits of Facebook or email users work. He found a pattern that is similar across multiple apps such as YouTube, Pinterest or Instagram. He calls it the HOOKED model and it has 4 phases (Trigger, Action, Variable reward and Investment).
There are two types of triggers, external and internal. We work with external triggers such as notifications. "Check out the new photos on Instagram." Notifications is the type of external trigger that Instagram uses.
How I did it
I created a pool system. There is a limited amount of tasks in a pool. The pool is shared among all workers, which creates competitiveness. If people are late, there may not be any tasks left for them. People are rewarded for every task they solve. If they are the first to arrive, they can earn more money because they will be able to solve more tasks. Pools are published irregularly and whenever they are published, users receive an email notification that acts as a trigger in the Hooked model.
With this system, we can manage the number of new leads coming to merchants based on their available capacity.
To increase the likelihood that the user will perform the intended action, the designer designs the action to be as simple as possible.
How I did it
Kamino will open a website that should contain the information we need. People are trying to find this information and copy it into the system. The reward they get by entering data into Kamino is visible in the system.
Variable rewards are one of the most effective tools that companies like Facebook use to build user habits. Research shows that dopamine levels rise when the brain expects a reward. Variable rewards on Facebook are likes and views.
How I did it
After entering the data (conference name and organizer) we are able to match it with our CRM. If it is a new potential customer, we give users a bonus task (for more money) in which they should fill in the contact details of the organiser.The staff do not know whether they will get new tasks at the basic rate or a bonus task for more money when they enter the data.
The investment can be in time, money or data. There is a lot of data that fills up Facebook, and that investment keeps you there.
How I did it
After some time, users will become more skilled and able to search and enter data more quickly. We have created a tier system in which users receive a percentage bonus. For example, at level 7, users get a 7% bonus on top of the base rate.This should provide long-term motivation to work.
I did 3 rounds of usability testing, one in person and two remotely. Several problems were found and subsequently fixed. 15 people were involved in the testing.
The autocorrection system wasn't working.
I expected users to correct themselves. I designed a checking system in which one task will be given to 2 users and if they enter the same data, the result will be considered correct and sent to CRM. But we found that we had a bigger problem. There is no way for unqualified students (temporary workers) to recognize whether a lead is suitable for us or not. So the only thing that would have prevented this would have been entering the wrong email. But you couldn't tell if an email/conference is appropriate or inappropriate for us. So I removed the autocorrect system and created a way for a qualified person from SlidesLive to evaluate a good match.
The product development was done by SlidesLive developers. I worked with them to create decision trees to enable further automation.
Před uvedením produktu na trh pracovali 3 lidé po dobu 1 měsíce v Kaminu. Z průměrné rychlosti těchto lidí jsem nastavil základní odměnu za jeden lead (0,3 $) tak, aby odpovídala průměrné hodinové sazbě v České republice v roce 2019. (130 Kč)
V beta testování jsem nenašel žádné zásadní problémy. Kamino bylo tedy připraveno ke spuštění.
We've built this platform to be scalable for over 1000 people. We needed to automate logins at this level.
Každý měsíc jsem pracoval s databází SQL, abych zjistil anomálie v datech. A abych našel způsob, jak optimalizovat algoritmus, také jsme vytvářel měsíční přehled.
The platform is used by 186 students and part-time employees.
The return on investment for the client was 259% after the first 6 months. Including staff fees and design and development costs, we were able to reduce the CPL (cost per lead) from $2.16 to $0.36.
