At Africa's Pocket, we are building a wealth management platform where you can learn, plan and act on your finances. One of the foundational aspects of building wealth is getting a grasp on how you spend your money. In line with this mission, we’ve been building a tool that analyses your financial statements (bank and mpesa) so you can understand your spending and plan accordingly to build your wealth - We’ve named it Coin Analysis.
At the beginning of every month, many of you spend time, sometimes hours, analyzing your statements to get insight into how you spent your money during the previous month.
This is how it normally goes:
Download all of your statements; bank and Mpesa accounts
Go through each transaction and categorize every. Single. One! This part is painful. How many times have you taken a boda boda or matatu in the last month?
Add all the transactions in each category to understand how much was spent across different categories.
Then use the results to make adjustments to your budget for the next month.
The process is long, manual, and exhausting. It’s no wonder so many of us struggle to stick to and review our budgets. If you’ve tried using an app to do this, you know that most of the apps are not customized for Kenyan budgets. You are forced to keep converting to USD or look at your numbers with a $ sign in front of them. Worst of all, they don't support the analysis of your M-Pesa and bank statements so you have to put in your information manually, a process that requires lots of discipline and will power.
The Process - How we Built this
The first step was to figure out a way of storing this information in the most secure way. For our customers to trust us with their financial information, We created our security based on the standards from the PCI Security Council Compliance, an international organization that maintains, evolves and promotes Payment Card Industry standards for the safety of cardholder data across the globe. Building our system based on the PCI standards ensured that we handled our user's information in the best way that has been agreed on by the bank industry around the world.
We then asked our users to tell us the banks that they would like the tool to support for analysis. The top five banks from the poll were KCB, NCBA, Standard Chartered, Equity, and Co-operative bank. The next step was to convert the statements from pdf format to a format that the system would be able to understand. We began with Mpesa because the data was readily available and later moved to the top five banks. We taught the system how to identify the various statements by telling it where to look and how to interpret them. We had to manually compare different statements, identify the patterns and then teach the system how we did it. The most challenging part was, the statements we had were not enough to cater to all possible scenarios thus making the system inaccurate. We came to you, our users, to seek help with getting more statements. Thankfully, you were happy to help and the system was able to identify the statements from the top banks - Asante (thank you)!
As a human being, you can look at the details of the transaction and identify which category it belongs to, but for computers, this is a little more complex. This step requires a lot of data that was not at our disposal when we first started out. We taught the system all we knew based on the data we had at that moment and built functionality for our users to help with the categorization. Some of you guys uploaded your statements and categorized the transactions the system was unable to identify. This then allowed Coin Analysis to recategorize transactions that had previously been misallocated. Thanks again!
The final step for the system was to analyze the categorized transaction and present the result to our user in a well-organized way. At first, we had over 20 categories, and presenting it on a doughnut chart was difficult, the data was hard to interpret due to the small size of the slices. Many of you complained that you couldn’t distinguish between the very close shades of green that resulted from this 😂
We had two options; reduce the number of categories or present the data in a horizontal bar graph. Once more, we asked you what you preferred. From the poll we conducted, 90% of the voters preferred a pie chart. So, we went on a mission to reduce the number of categories to twelve, we changed the slicer colors, and organized the chart so it’s easy for you to understand.
Introducing Coin Analysis
Coin Analysis will analyze and organize your financial transactions, allowing you to see patterns and identify opportunities to save more. You can combine statements from different banks and mobile money wallets, analyze them, and visualize the results in one dashboard so you have a full overview of your entire financial life.
We’re grateful for the support we've been getting from you, our community. As we continue to develop Coin Analysis our decisions about what to build are largely driven by you. We hope to make the process of understanding where your money is going as easy as clicking two buttons, so you can spend more time enjoying your money than you do analyzing it.
We've been hosting monthly beta tests with a small group of you and are continuously collecting feedback to make coin analysis serve our larger user base better. If you are interested in joining our beta group, please join the waitlist here. You can also contribute to the feature list by making your requests here. We look forward to analyzing all of your coins soon and getting you one step closer to building generational wealth.