This contribution is a translation from the German original authored by Lukasz Glowacki:
“Customer acquisition with algorithmic elegance”: photono combines state of the art data analysis algorithms and artificial intelligence with a diverse pool of data sources to generate new knowledge.
So that you know in advance, where your sales pitch will really generate new revenue.
Editor: Hello Justin, hello Hauke, can you tell me how the idea for photono emerged?
Justin: Ha, that’s a good question and the detailed answer would be almost as long as the story in “How I met your mother”. After all, photono is actually building on years of experience and know-how from various projects and initiatives in the EWE DataLab and the major energy transition project enera. So the idea for photono has been maturing for a long time and the implementation is based on a lot of development work which was done in the run-up.
Hauke: When Justin told me about his vision for photono in the fall of 2019, while searching for comrades-in-arms, I was actually quickly excited about the idea and the approach, because on the one hand, the information retrieval from aerial images supported by artificial intelligence is technologically exciting and groundbreaking and is not yet very well developed in our industry. On the other hand, in various projects we have repeatedly experienced the great pain that exists in the area of efficient, precisely tailored customer acquisition. And I believe that our solution can provide very good support in this domain.
Justin: We actually started out very shirt-sleeved with a simple poster, with which we searched for internal customers in the EWE Group and with which we pitched the idea in the EWE Digital Factory for implementation. Of course a lot has changed since then, but the core of the value proposition to our customers has remained the same:
Editor: Before we get to the customer benefits, what does implementation in the EWE Digital Factory mean?
Justin: The EWE Digital Factory is part of the business unit innovation within EWE and, following the successful pitch, offers us the opportunity to develop photono like a start-up and align it with the market. In particular, it gives us the resources we need to develop an idea into a functioning product. At the same time, we are part of the EWE Group, benefit from the expert knowledge available here in a wide variety of areas and have already successfully implemented our solutions in various parts of the Group.
Editor: That sounds like an exciting win-win situation. Which brings us to the topic: What benefits does photono offer its customers?
Hauke: With photono, we support our customers in the areas of lead generation, product placement, and perspectively also risk minimization. In lead generation, for example, the photono solution enables us to carry out precisely tailored customer segmentation and, thanks to its modular structure, to identify potential new customers quickly. In other words, by using our algorithms, we reduce customer acquisition costs on the one hand, or optimize customer lifetime value on the other, depending on customer requirements.
Justin: In addition to this, we offer our customers support in product placement. With the photono solution we enable targeted advertising that helps everyone: The advertiser has less effort and the advertised person optimally only receives what he is really interested in. A further advantage are individual offers, e.g. towards commercial businesses. Here, for example, an automated amortization estimate can be made in advance of the customer approach for certain products such as photovoltaics large roofs. This is a real advantage, because as a key account manager you save yourself a few iterations with potential customers and can immediately come up with the best argument: that it is worth it! And of course we also calculate affinity scores for our customers on a product-specific basis, which makes the acquisition process much more efficient.
Hauke: The third area we are currently preparing is risk minimization. Here we want to use our AI algorithms and analyses to improve business forecasts and thus also enable business case validations. With our feature analyses we can, for example, quantitatively validate business case hypotheses for campaigns or products relatively easily – and thus minimize the risk of business failure.
Editor: That sounds very helpful, but also pretty comprehensive. What specific solutions does photono currently offer its customers?
Justin: We already offer various data services in production and benefit from the modular structure of our solution. For the time being, the data services are being offered under the generic term AI-supported geomarketing. Our main use cases at the moment feature the linking of the modules trade directory 4.0, photovoltaics database and customized aerial image analysis using AI in combination with our geo-databases. Of course you can also book individual modules.
Hauke: Or combine these modules with others, for example the module for roof area potential for photovoltaics, which we are currently testing. And all these modules then lead to affinity analyses for our customers, which answer specific business questions: e.g. for which commercial customers large-area PV systems or battery storage could be useful, where the affinity for electromobility is particularly high, and many more. You can find our current product range at https://photono.io/en/startseite-english/#portfolio.
Editor: Can you give a short overview of how this works technically?
Justin: As I said, the linchpin of our solution is the modular photono architecture. In the field of aerial image analysis we also use state-of-the-art AI algorithms, so-called Deep Neural Networks, which are trained with generic image material and which we then teach to identify the features and attributes of interest to us for our analyses. I think I can confidently say, that we are now really good at this.
Hauke: In addition, we benefit enormously from the great IT expertise, especially with regard to geodata, in the photono team and in the EWE DataLab, and also from the continuous exchange with the research institutions located in and around Oldenburg. In this way we are always getting new impetus for our solution, whose core technologies are still very new and are also developing very quickly.
Editor: In what way does photono differ from other solutions on the market?
Justin: The idea of using satellite images to obtain information is actually not entirely new. The first thoughts on this were already made in 2007: https://newatlas.com/satellite-imagery-used-for-sales-lead-generation/8063/. In my opinion this is not entirely surprising, because the potential of the technology used is very large for all kinds of industries and applications. Today the technology is much more advanced and we at photono actually use cutting-edge algorithms to achieve very high quality in our analyses.
Hauke: Otherwise, as far as we know, there are not very many other players with a similarly strong energy and telecommunications industry background as we do have. And we have already shown in the first projects that we as the photono team can implement customer requirements individually, quickly and safely. Security, especially with regard to data protection, is of course also a very important and essential aspect of our solution, which we are working on.
Editor: Okay. I understand you are still a very young project. So what’s going to happen next with photono?
Justin: At the moment we are very optimistic. We have several target groups in mind for which photono could be helpful. And we have already had several encouraging discussions with potentially interested customers in recent weeks. If you feel addressed and are interested in seeing our Minimum Viable Product in a live demo, please feel free to contact us anytime at email@example.com.
Hauke: Or register on the website for one of the webinars on the subject, which we hold regularly: https://www.photono.io/web-seminar
The next one will take place after the summer break at the beginning of September and we are of course hoping for many interested people who we want to convince of our approach “customer acquisition with algorithmic elegance”!
Editor: Thank you very much for the interview!
Justin: We thank you in the name of the whole photono team!
Dr. Justin Heinermann – Founder & Data Science
Dr. Hauke Held – Data Science & Business Development