This privacy notice discloses the privacy practices for InFlow Machine Learning Inc. This privacy notice applies solely to information collected by this website. It will notify you of the following:
We are the sole owners of the information collected on this site. We only have access to/collect information that you voluntarily give us via email or other direct contact from you. We will not sell or rent this information to anyone.
We will use your information to respond to you, regarding the reason you contacted us. We will not share your information with any third party outside of our organization, other than as necessary to fulfil your request, e.g. to ship an order.
You may opt out of any future contacts from us at any time. You can do the following at any time by contacting us via the email address or phone number given on our website:
We take precautions to protect your information. When you submit sensitive information via the website, your information is protected both online and offline.
Wherever we collect sensitive information (such as credit card data), that information is encrypted and transmitted to us in a secure way. You can verify this by looking for a lock icon in the address bar and looking for "https" at the beginning of the address of the Web page.
While we use encryption to protect sensitive information transmitted online, we also protect your information offline. Only employees who need the information to perform a specific job (for example, billing or customer service) are granted access to personally identifiable information. The computers/servers in which we store personally identifiable information are kept in a secure environment.
In a world where we're told data is the new oil it's never been easier go up in flames when trying to find the meaning of it all. This post focuses on strategies to derive insights from mountain-high haystacks of data.
At InFlow we pride ourselves on our ability to extract exorbitant amounts of information from a single video camera, but when it comes to deriving insights we know just how hard it can be to find that proverbial needle in the haystack. This article will highlight ways to think about and analyzing your organization’s data with the objective of pulling out what really matters.
Let's just start by saying there isn't one. Data can help us tell a story or it can help us fabricate one to our liking. Your data is not a treasure map that will lead you to a 7% increase in profits or 147 new monthly active users, instead it should be treated as a series of interconnected clues that lead to a realization. There are plenty of data analysis tools and AI systems designed to collect, store and sift through mass amounts of information and extract patterns or trends, but no matter how much data you have you will never truly have the full picture. Whatever tools you use, including InFlow, it’s always important to build real world context into every trend or anomaly.
There's a powerful difference between a fact and an insight. Facts are simply trends or anomalies in your data. A fact becomes an insight when it can be explained and an insight becomes valuable when it can be leveraged. Here's an example, when analyzing pedestrian traffic in a retail store we find that on average people enter the store and go left 70% of the time. Until we better understand why this happens it is simply an interesting fact. For now then the best action we can take is to position some staff on the left side of the store and maybe move some displays or products in that area as a way to ensure the majority of people entering the store see those items.
The data we first collected and the adjustment we make is not enough. We can't just stop at this step. To really leverage this data, this fact, we have to derive reason and meaning. The most powerful insights always come from the most simple questions. I won't steal credit from Amanda Lang's The Power of Why (which you should read if you haven't already) but ‘why’ is precisely the question we need ask and answer to turn our retail fact into a powerful insight. Answering the question 'Why?' is the hard part!
So before we jump to any conclusions, let's take a step back and add some real world context to this fact. To do so we'll need to answer a few secondary questions:
The answers to our secondary questions provide us with a broader context. We learn that products seem to sell better on the right side of the store despite the traffic being driven to the left. We discover that only 30% of people who enter do a full loop around the store. Finally, we note that our clearance display is positioned immediately left of the entrance, resulting in another question -- might the placement of that display be a factor?
Knowing what we think we know, the next step is to test and adjust. We think that maybe the clearance display is driving people to the left when they enter the store, but we don't know why products are selling better on the right side. To test this arrangement, we reposition the clearance display to the right side of the store. Immediately, we observe more traffic to the right but a sales dip in all products positioned on the right side of the wall. Great. We've validated that the clearance display is controlling the direction of traffic upon entering the store, but we have another question: why do the sales of our co-located products dip? We quickly replace the clearance display with featured items from the right side of the store and see that our traffic analytics show less people are entering the store but we have a much higher conversion rate on each customer. There are probably more questions and adjustments to be made to get all the parts working at an optimal level for the retailer, but let’s move to insights.
At this point, it's clear that the clearance display has a negative impact on sales on the products around it and it plays a critical role in driving traffic into and through the store. We can assume, to some extent and with some confidence, that the clearance display creates a sticker shock effect on all the items in close proximity making those items seem expensive in comparison thereby reducing their sales. We've now turned our fact into an insight, and can leverage this understanding to improve performance. We now make the decision to move the clearance display to the rear of the store with the other lower priced items. Our data now reveals that people the number of people that go left or right when entering is mow close to an even split. Even better, 80% of the people do a full loop of the store.
It's incredibly easy to get bogged down by facts when analyzing the details and trends in your data. We can forget or quickly put aside the inherent connectedness of data. Going beyond the data, beyond the point of fact is a ‘must’ do, and to do that that we must ask 'Why?'. We started our exploration to identify insights by using one particular fact -- 70% of people entering a store went left, but we could just have easily started with another fact -- products on the right side of the store sell better than those on the left. Both appear to unrelated, and had we not dug deeper, applied real-world context to the fact we would not have discovered the true opportunity for improvement.
A fact, when coupled with the mindset of ‘why’, results in an insight – a needle in the data haystack - that positions you to leverage your organization’s true potential.