One of the mistakes people make is assuming that data are infallible. When some individuals want to prove the accuracy of their assertions, they back them up with related data and expect everyone else to accept their argument. Of course, it is always good to use facts and figures as a buffer for arguments but that doesn’t mean that what is said is true. The truth is, “data” is not always right. This is one thing you should always bear in mind each time you pull up published information or data as the basis for your argument.

You see, what is written on paper or published on the internet is not the problem but the processes it passed through before it was published. If you had noticed, while you were in school (or even right now in your office), you are always expected to use different sources for writing an assignment, a proposal, a report, or a project. The people that tell you to do so are not out to stress you but to make sure that the imperfection in the data you are using is reduced. If you pay attention to the differences in the data published about a particular phenomenon, you will discover the discrepancies in them. Of course, ordinarily, factors, such differences in age, gender, class, race, religion, cultural orientation, financial status, social environment, and so on, affect data. But data collected from the same group of persons within the same time frame and by the same persons (or maybe different persons) can come out differently. This is why you shouldn’t always “swear” that what is published is the absolute truth because you think it is objective.

As mentioned earlier, sometimes, the same set of data can be collected from a group of respondents but they will not give the same result. The factors that can cause these discrepancies can come from the data collection method, the data collector, and the respondents themselves.

The method used in collecting data affects what is gathered. For instance, a questionnaire is supposed to be the most unbiased method but then, it makes respondents self-conscious and less willing to reveal sensitive information. People easily lie on questionnaires. Then, the interview is another avenue that allows respondents to be studied objectively. But it is not all information that can be gathered through interviews. Besides, respondents can also lie through this method. Finally, the observation method is usually the best when the data collector wants to get unbiased data from respondents. However, this method works better when the observation is carried out over a long period and the data collector is not biased.

Data collectors also contribute towards making data fallible. People sit down in their houses to falsify information based on “what they see happening around them.” They assume for the respondents and, so, supply data based on what they think and not on what is real. Some collectors also manipulate respondents into supplying the type of information that suits their (that is collectors’) narratives. There are also cases of misrepresentations, where data collectors select respondents that are either unsuitable for the research or those that will supply the desired information. All these and more are reasons you should be careful with how you rely on published data.

There are so many factors that can make respondents supply different information at different times. The first one is emotion and other physical conditions, such as illness, hunger, and the rest of them. Then, there are respondents that deliberately supply inaccurate information or even refuse to supply any. Some misunderstand what was demanded from them and many provide information to impress the data collector.

You know, the reasons mentioned above are part of why the use of secondary data is not always encouraged in any research project. Unless you are the one that collected the data, you cannot boldly use them to arrive at a conclusion. This is just to say that you should take all these figures you see out there with a pinch of salt because not all of them are facts.

As the world battles with the identity crisis, struggling through the challenge to differentiate one face from the other through proof of identity that involves an estimated one billion people who lack official proof of identity, it beckons a huge task.

Though taking on the proof of identity challenge presents a mammoth task, especially in Sub-Saharan Africa and South Asia, where 81% of the involved 1 billion people live, according to World Bank Group’s report, Smile Identity is taking on it with uncommon determination.

The company provides ID verification and KYC compliance for African faces and identities, and it is getting the needed funding. On Thursday, the company announced that it has closed a $7 million Series A funding that new and existing investors participated in.

Costanoa Ventures co-led the investment with pan-African venture firm CRE Venture Capital. Other investors that participated include VCs like LocalGlobe, Intercept Ventures, Future Africa and unnamed angel investors. Existing investors, including Khosla Impact, ValueStream Ventures, Beta Ventures, 500 Startups and Story Ventures, also participated.

TechCrunch reported on the company’s journey since 2017, moving from an idea to a working solution to the world’s identity crisis.

Mark Straub founded Smile Identity with William Bares in 2017. As an investor, Straub took regular trips between Mumbai and Nairobi, spending up to a decade in the former. In the early 2000s, it was incredibly frustrating to get SIM cards or provide identity checks in India. But things began to change in 2009 when the India Stack was set in motion. The stack is a unified software platform via APIs that provide governments, businesses, startups and developers with seamless identification, verification and authentication processes across multiple industries.

People could easily do things like KYC to sign a document digitally, to share or federate documents across a company or school, to prove a birth certificate, or to prove an educational credential, he said to TechCrunch. It really unleashed this wave of innovation, entrepreneurship activity and fundamentally changed financial services and eventually shared economy services in India.

While India greatly advanced with this effort, Africa has largely played catch up. Most businesses on the continent still find it daunting to onboard new customers, salespeople or employees without spending lots of time and energy on people, processes and paperwork.

I’d seen for 10 or 15 years how long Indians had been waiting and how many problems this set of new national ID combined with these different software protocols solved. I saw how much friction the Stack removed from people’s lives. And I thought if only there was an Africa stack.

Straub said he began brainstorming with some entrepreneurs in Nairobi and figured that if an African Stack was to be built, three factors needed to come into play: cost, government independence and spot-on technology. It also happened that at that time, open-source face recognition algorithms on a wide range of smartphones had flooded the market and advanced the state of verification dramatically.

Leveraging camera technology, Straub and his partners started Smile Identity and developed de-biased face recognition that was ultimately more accurate for Africans than the base level open-source algorithms published in U.S. colleges.

We were able to combine that over time by matching selfies against either identity documents or photos on file at ID issuing authorities. And it was really the combination of those two technologies – the face recognition and all the liveness checks and, and anti-fraud checks that go with it, combined with verifying for the source of truth.

Four years on, Smile Identity is now present across six markets in Africa: Nigeria, Kenya, South Africa, Ghana, Rwanda and Uganda. It covers more than 250 million identities and verification for 15 different ID types while performing over one million identity checks every month.

Its software is used in banking, fintech, ride-sharing, worker verification, public social welfare programs and telecommunications. Smile Identity says it has about 80 customers who are charged on a per-query basis. Some include payments companies like Paystack, Paga and Chipper Cash; neobanks like Kuda and Umba; traditional banks like Stanbic IBTC; cryptocurrency exchanges like Binance, Luno and Paxful; and supply-chain businesses like Twiga.

Finding product-market fit took a while for Smile Identity at first. There was little or no playbook to follow, so the company had to figure out the right mixture of features and pricing favourable for African markets, smartphones and the internet.

We have had to build SDKs and mobile wrappers that work for Android and iOS but also things like React Native and Flutter and make them work on mobile web browsers. Almost every one of these solutions we came up with was because we hit some friction point with customers that forced us to innovate. A lot of those were either device issues that we ran into or low internet connection or low or no internet connection conditions.

But with continuous iteration and working with different clients, Smile Identity is on track for another level of growth. In the past couple of months, there have been numerous talks on fintechs and regulatory requirements ranging from local data protection laws and consent requirements to privacy policies and data impact assessments. These requirements are hard to keep up with, and Straub says Smile Identity is working on templates its clients can use to navigate them.