The future looks bright
Lots of talk about artificial intelligence these days, right? Pretty much every tech company on the planet is boasting their advances in machine learning and some of it – let’s be honest – is more than a little obscure, making it kind of hard to see what the point is.
In the TASK Lab, we’ve been working on our vision for AI with a focus on its practical application – making sure that we’re developing tools and product features that not only take advantage of new technologies, but will deliver real, discernible and measurable benefit to our customers.
Not familiar with TASK Lab? Well, that’s where all the good stuff happens. Its where our engineers and developers come up with cool new features and incorporate them into our existing solution.
Using the TASK Lab as a starting point means the team can test new features and functionality in a monitored environment designed to replicate the same demanding conditions found in our customer sites. So, by the time you see something new in the latest release, you know it’s been tested and proven to work in the real world. That practice of continual improvement and innovation is part of our promise to our customers.
But, it’s not all about today, our TASK Lab gurus have got an eye focused on the future as well. That’s why they’re working on a staged development program and building algorithms that utilise the vast amount of data captured by the TASK transaction management engine every day to extraordinary effect.
Just because you haven’t seen it yet, that doesn’t mean we aren’t working on it. We’ve put efforts into developing our basket analysis algorithm because we know that understanding the relationship between purchased items is the key to increasing profitability.
We’re using the transactional data we already capture to learn associations between products so we can dynamically incorporate that info back into our sales engine. That means we can clearly identify frequently correlated items and build those into our upsell/cross-sell engine, better understand common combinations to improve product placement within menus and dynamically create new combos that match natural pairings already made by customers. The upshot is smarter, more capable kiosks, POS systems and online ordering and you’ll see it in near-future releases.
We’re also pretty excited about our computer vision project, which captures visual information at kiosks and POS and allows us to unlock demographic user insights without the need for a loyalty or member scan, swipe or cell phone number. That information is stored at a transaction level and enables us to learn more about the current customer base including demographic strengths, weaknesses and opportunities. It gives us greater insight into product preferences at a demographic level and to explore targeted upsells and cross-sells designed to boost redemption rates and increase revenue. Of course, this functionality already exists in the TASK solution, but we think that facial recognition technology offers a faster, more efficient path to demographic-based targeted offering creation.
A little further down the track, you’ll see the results of our targeted suggestive selling development program. Building off the back of our basket analysis algorithm, the TASK system will offer targeted sells based on the probability of uptake, ensuring recommended items are relevant to the current user based on demographic data and redemption probability.
The system analyses the state of a user’s populated cart both during order flows and at the point of checkout to determine upsell and cross-sell items and, because the algorithm is learning with every transaction, accuracy levels will continue to climb over time.
Our loyalty module is getting some AI love as well, while we work on customer segmentation and propensity modelling. We’re enhancing our member segmentation tool with unsupervised learning techniques that allow discovery of hidden user groups that share similar behavioural patterns. Those clusters are then dissected and analysed, offering more detailed customer information based on frequency and recency of visit, product preferences, demographics or a mixture of all.
This means your marketing team can target customer communications at times and in ways that have a higher likelihood of being well received. Dynamic offers can be generated for each user, giving customers a personalised experience through tailored product offerings based on their unique preferences.
We’re excited about the work we’re doing, implementing our AI program and teaching algorithms how to better utilise the copious amounts of data our transaction management engine already captures. We love being at the forefront of what we do, because it means we can continue to deliver a better solution, while you and your customers reap the benefits that our AI program provides. From where we stand, the future looks bright.