Here's what you'll learn:
Data volumes are increasing continuously, but the challenge is to use them profitably for the company. The successful application of Big Data is enabled by its targeted, ongoing generation, processing, analysis and optimization. Through Data-driven Excellence, the impact of all digital activities on a company's key KPIs and business model can be measured and thus managed for all core digital areas.
In addition to concrete procedural options and best practice examples, we show for which application areas data management is relevant for Data-driven Excellence, how companies can break down data silos and use Big Data for themselves, which data strategy should then be pursued, and which fields of action must be addressed in order to successfully establish Data-driven Excellence in companies.
Watch online now (German only):
Download the presentation now (German only):
Transcript of the Webinar Data-driven Excellence
Welcome & Introduction
Angela Meyer: Welcome to our diva-e webinar Data-driven Excellence. Today, you'll learn from our diva-e experts Albert Brenner and Jan Stöckel how you can gain relevant insights about customers and their digital behaviour from data silos. My name is Angela Meyer, and I work in the diva-e-marketing team. And I'm in charge of our events and webinars, and I'm your presenter today—Jan and Albert, a few words about you.
Albert Brenner: I'd be happy to, Angela. Albert Brenner, I am one of the co-founders of diva-e, responsible for strategy and data consulting. I'm a native of the digital world. It has to be said. I founded the first websites in 97, and since then, I've always been very committed and happy to support companies in their digital transformation.
Jan Stöckel: Hello, my name is Jan Stöckel. I've been with diva-e for ten years now. I've also been involved with website and e-commerce topics since the 2000s. I'm currently responsible for customer experience and data as part of the consulting team. I support clients in everything from digital strategy to the design and development of e-commerce and customer platforms. And as we will see in a moment, data plays an essential role in this.
Angela Meyer: Then we will now start with the presentation. Albert, I'll hand over the broadcast rights to you. And I hope the participants enjoy listening.
Albert Brenner: Yes, thank you, Angela, for the intro. And I'd like to welcome everyone again. I'm pleased that Jan and I will be able to tell you something about the path to data excellence in the next 45 minutes or so. In other words, how can I use data in a structured and intelligent way to generate relevant business success? Jan and I will now take turns in the webinar, and we will first start with the question of what you will take away from this webinar.
From our point of view, the first and most important thing is that we've removed the most significant hurdles from the topic. Sorry, now I have here-. That we will be able to point out the most critical limitations of the issue, and in principle, go into what can be the blocks of obstacles, so to speak, in the topic of Data Excellence. Where is the most significant potential for companies in data? We'll talk about the holy grail of AI, and we'll also try to de-mystify the topic a bit. We will discuss what questions you should ask yourself to achieve digital excellence, especially data-driven excellence. And what to expect after developing and implementing a data strategy. What do I get out of it? What are the benefits? And what is the best approach ultimately to achieve data-driven marketing and sales? And finally, and this is undoubtedly a critical point, how do you get started quickly? That the topic of data-driven excellence gets off the ground, so to speak, and does not require a year-long preparation phase. Before we get into these solution topics, so to speak, I would like to talk about the issue of challenges. And concerning challenges, what are the main drivers for data excellence?
Challenges and critical drivers for data excellence
And in principle, we see two main drivers here. One is the change in customer requirements, driven by digitization. The customer experience that I have gained with Amazon, Google, Apple, Facebook, and the like is what I expect when working with other companies. An expectation level is set. And these companies achieve customer experience. Thus the competitive advantage they have today, mainly through the intelligent collection and use of data. That's sort of one driver for that. And the other topic, the other driver, is that digitization has something to do with systems, with tools. And very often, in our projects, we find that companies simply have an inevitable fragmentation of systems keyword data silos. And that, of course, also leads to the complexity of using data profitably. Let's take another closer look. What are the main problems? Customer expectations have already been briefly addressed in principle.
The expectation is that I will not be addressed as an anonymous customer but will be provided with highly personalized information and offers relevant to my situation. This is also consistent across all customer contact points. The second is the explosion of these contact points. We used to interact with customers via two or three channels. We now use five, six and seven intensively. So it's the omnichannel approach. And each of these channels naturally has its systems and data silos behind it. And how do I manage to create a consistent customer experience across these channels? In the data, in the system, different data have, let's say, different standardizations, different structures, different qualities, which are also not all so directly available. This also creates complexity. And ultimately, the big question is, what do I want to achieve with the data?
The GDPR describes reduced and targeted data collection. I can't just collect data wildly. I have to do it on a fundamental level. And this raises the question of the fields of application for the data? What can I ultimately achieve with data in my business model, in my structures, in my go-to-market? And that is the question of data strategy. And then ultimately also about the data management infrastructure. From our point of view, data is one of the main topics of digital transformation. That means, conversely, no digital change without data-driven excellence. And that's what we want to take a closer look at now.
No digital transformation without data-driven excellence
So, where are the levers at the end of the day? And in terms of levers, we see in principle that data leads to competitive advantages. But the question is, at which point, in which area can I ultimately achieve competitive advantages? And what we see is, first of all, that the most powerful lever for data excellence is in marketing and sales. That's where the customer contact interface, customer care, so to speak, all dialogues and interaction processes with the customer are. That is also the most significant potential. Everything that relates to pricing management and promotion, so ultimately campaigns, dialogue channels. The third thing is, what do I eventually offer the customer? Next Best Offer is a keyword that you hear repeatedly. Next Best Action. That means an individualized offer for the customer in his specific situation. That can be B2B, but it can also be B2C. It doesn't matter.
Ultimately, it's about which products and services I offer the customer at which point in the interaction process. And the topic of sales, customer acquisition, lead generation, lead qualification. In other words, where do I focus my Euros on customer acquisition? These are the most significant levers. Of course, there is also much potential in the supply chain, in production, IoT, networked production, proactive maintenance. But in total, it's at least a tad smaller than in marketing and sales. And then in post-sales, there's data potential there as well, of course, but that's the most miniature lever. The third thing I would like to point out is that there is always much talk about AI. The hint at the point, in 95 per cent, at least in 95 per cent of the cases where AI is talked about, we talk about sometimes relatively trivial statistical models. These can also be more intelligent models, such as data mining, neural networks and machine learning. I don't want to go that deep into it. But we always talk about artificial intelligence. You can see here that by far the most significant leverage, a tremendous potential, actually has nothing to do with artificial intelligence but is simply partly quite average statistics. And there, the challenge is instead to bring together the correct data and build a suitable statistical model. Now, in the next step, and here I would like to hand over to Jan, we want to look at these components and modules. Especially in the area where the most significant potential lies for the intelligent use of data in marketing and sales. And Jan will present us with a construction kit.
Building Block Data-driven Excellence
Jan Stöckel: Yes, thank you Albert. Once again, welcome from me to the newly connected participants. If we look at the levers, as Albert already said, it is essentially the area of marketing and sales that stands out here and where we see a great deal of savings potential, particularly in the area of acquisition and optimization of the purchasing process. And as Albert already said, we have a toolbox where we look together with you where there is a need for optimization. And the first crucial area is to look at the customer contact points across the individual channels. We call it the customer journey. And, of course, identify these touchpoints and provide them with corresponding KPIs, i.e. key performance indices. In other words, we measure what happens at various contact points, where it is digitally possible, using metrics, what the user does, how he behaves, and so on. This enables us to conclude activities. This means, for example, that we can run marketing activities by running SEA campaigns and then look at how many clicks on the campaign, how many come into the store, and how many convert or cancel. Such evaluations are then standard and give us an excellent insight into the activity performance. This means that campaigns can be evaluated very well using such measures. In addition to the customer's behaviour, we naturally add things like what the campaign costs us in terms of budget and can then, of course, plan the campaigns with more budget that are more successful.
SEA is then pitted against display advertising and the like, and conversion rates or other KPIs are used to determine the most successful, which has the best cost-turnover ratio and is then naturally allocated better or higher than others. In other words, we optimize directly at the touchpoints. And in this way, we achieve what is known as attribution optimization. This means that the channels' role and effect in their sequence are closely monitored and supported in an optimized manner. It's also about retargeting, looking at which path the user takes and which direction is the one with the highest completion rate or success rate. And then, of course, when do you provide the customer with certain information, or how do you give him information? We can deduce where to send data again if we have contact data here. Or in the direction of personalization, which content is displayed to the customer in the store, for example, or which product ranges interest him, and we give him a discount code.
Of course, the whole thing also does in terms of costs and sales is that you get a transparent view of the entire customer journey. In the end, you get your colleagues to realize that it's not just the store manager who runs the store but that the customer is viewed as a whole. And his customer journey from the first interest to the conclusion and after-sale. And through these measures in marketing sales, there are also numerous in other areas, but that would now lead too far to detail, you develop optimized marketing campaigns as we have already seen. And in particular, we get insights into how we can expand the business model. How can we digitize products to receive additional services digitally, become digital themselves, or provide supplementary services as already mentioned? Of course, you get transparency about customer channels, and you have much more insight into your users and how they behave.
Companies like Facebook, for example, also explicitly pass on information on age structures, behaviour and the like, in some cases anonymously. And with this kind of information, you can, of course, draw conclusions about what is happening at all, which target group is responding to which activities, which target group is particularly active in which channels. And you can then carry out very detailed target group clustering and play out the campaigns accordingly. Furthermore, in the projects where we use marketing and sales analytics, a preliminary process is usually necessary to look at the data from the past and then create forecasts. We evaluate which channels will generate traffic with probability and visitor numbers based on these forecasts with concrete KPIs. And then still derived via historical data. What are typical conversion rates, and in the end, standard shopping carts and the like, excellent planning can be carried out for a company. Based on this planning and this prediction, we can start a forecast, which is always compared in parallel with the actual state. And you have excellent transparency about what activities are currently running. Are they running as planned, or do you have to add to them if necessary and activate additional newsletters, for example, to perhaps still achieve the sales in the week of the quarter. So everything is also related to periods. And in the end, the entire customer communication is optimized and customer-oriented.
As I said, the whole thing is fact-based, which means that we don't make decisions based on gut feeling but on concrete KPIs that are defined together with you. And, of course, we draw on our extensive experience from many commerce and other projects. And a structured approach is a matter of course. But we will come back to that in detail in a moment. And I want to show you again on the next slide that when you use systems like this, it also moves the company forward.
Because it has been shown that two-thirds of the leading organizations say that they can ultimately make definitive statements about the business using such methodologies, and they don't just do it on gut feeling. That is certainly a statement that you also deal with your company in the current world, where we are confronted with data. And 70 per cent of the marketing experts also say that their companies, or in their companies, the decisions are really made on data, and no longer purely from forecasts that are perhaps based on average values or generally valid numbers—but based on their analyses. And data-driven organizations usually have three times more frequent decision-making processes. It is much more effective to work with data in this way than perhaps through old-fashioned methods.
Factors influencing data-driven excellence
And the fact that it's possible to work with data at all these days is, of course, due to technological advances. As Albert already said, it is a significant part of digitization. And today's systems and modules, I would say the extensions to services that cloud products and e-commerce systems, CMS platforms, and analytics and online marketing tools bring, now enable a straightforward setup. We see that right away in these kinds of analytics capabilities. The data is no longer in silos but can be consumed and enriched more easily. If you want to continue to deal with the topic, say yes to the issue of data strategy, data is an essential topic for my company. You must, of course, deal with specific questions. And we have brought these along for you today. So the eight most important questions that you should ask yourself from an entrepreneur's point of view to start with the topic. Once you have answered them, you can certainly enter a direct dialogue. And the first one is, of course, how can data-driven processes or characteristics influence decisions in the company in the future? And how can conclusions be derived from that as well? And that ultimately describes the area of Use cases very well. One deals with how one can achieve an effect operationally with data. This includes examining the surrounding processes in detail and looking at the bottom line: do I already have data at this point, and where does data leak out? Where is it located? Is it in silos, or is it already consolidated? And how can I optimize the process? In other words, what is my specific goal, what do I want to achieve? Do I want better or higher customer satisfaction, or do I want more sales? And these are the questions you should ask yourself in this process. And then, together with a data strategy, formulate the use case in detail.
That means that today we are mainly talking about marketing and sales processes. That means prioritizing concrete effects. You have indeed identified numerous use cases, and then you have to think about which one is perhaps the use case that we should launch first. In other words, we should look again at the requirements. That is, what are the external and internal requirements for the process? Albert has already mentioned that. Is that perhaps also something with DSGVO? If you address customers directly, of course, you need double opt-in to work there possibly and the like. This is explained in detail in the requirements management. And then, we look at the current situation. That is, we look at the company's maturity, where are you, what tools are you working with? Do you perhaps already have a system onboard that can be used to implement data-driven processes, perform evaluations, or maybe build dashboards directly? Or does it all have to be set up from scratch? Then it depends on your preferences as well. Do you work on-premise, or do you already have cloud applications? And we end up deciding on a specific technology, so we work with third-party tools. We'll introduce you to those in a moment. Yes, then it's about building out the data-driven capabilities in the company; usually, there don't exist that many people there yet who are intensively working on that. You need data analysts, which is now a rare commodity on the market. And colleagues who understand how to deal with it, know how to read data, and, more importantly, deal with it. How is it visualized to make the information available to the business? Data governance is also an important topic. How is the data handled, what authorizations are there in the company, and who can access where? And as I said, DSGVO and the like also play a role.
And the last important question. Is there a structure or, in the final analysis, an initiative for digital transformation in the company, affiliated with and accompanied by the data-driven enterprise, to ultimately create the prerequisite for you to work with the data at all? It is not a different project that is considered separately. If you have dealt with the questions, you are well prepared to start such a project. And we are looking forward to your answers. Now I would like to hand it over to Albert again, who will briefly introduce best practices and the possibilities to evaluate data in the company.
Albert Brenner: Thank you very much, Jan. Before we do that, we had thought in the webinar- the dialogue capability is limited, of course, but we felt at least times a short impulse from your side from the audience. Angela, for that, we prepared a question for the group.
Angela Meyer: Exactly. Thank you, Jan and Albert. And we have prepared a question for participants to determine whether they are already using Big Data to gain relevant insights about your customers and their digital behaviour. I'm going to start the survey here. And you have time to answer this question now. And afterwards, we will analyze the results together in the Q&A session. We'll give you a few seconds. Exactly. We will then look at your answers together in the Q&A session. That will be interesting. So, more than half of you have already voted. And I would like to close the poll in ten seconds. And then, Albert can continue with the presentation, and we'll look at the survey results later. Yes.
Data analytics in the enterprise
Albert Brenner: Cool, great. Thanks, Angela. We'll get to the survey results right now. Now, data analytic voodoo is happening in the background to look at that right then. Speaking of a curse, how do other successful companies do it? We have brought along four examples to simply show a range of how other companies deal with the topic of data analytics. We accompany many companies, and there are five topics or five skills, I would say, that distinguish companies that are successful with the use of data. One is the fundamental trust in data. That means that if I ultimately say I want to automate a process based on data, or if I want to make decisions based on data, then, of course, humans can still eventually make the final decision. But I also have to trust the collected and processed data to support my choices. And for that, the data must also be trustworthy.
This means that data quality, consistency, and completeness must also be a given. The second is a fundamental commitment to seeing data as a strategic asset. It's not for nothing that people say that data is like looking for gold, and then they look for the nuggets in the data, the little pieces of gold in the data, employing analytical processes. In other words, data is the new oil, the new raw material of the digital era. In other words, I need a specific commitment that also leads to investments, for example, that also leads to resource allocation, to see data as a strategic asset. I need the necessary talent. Jan has already mentioned that data scientists are currently in demand on the market. And the data is not created by me keeping it, or the data treasure is not created by me holding the data for myself, so to speak, be it the employee, be it the business unit. It's about bringing data together. And it's the consolidation that really creates the necessary basis for deriving intelligent analyses and, ultimately, the resulting insights. In other words, the organization's mindset is also concerned with data. Now let's look at four companies that we have already promised. How do they do it, which use cases do they serve? And we have now singled out OSRAM.
I think the company is well known. As part of their digital strategy, they have identified data-driven decision making in marketing and sales as one of their areas of action. And they have proceeded in such a way that they have said which goals, which business goals must be achieved by marketing at the highest level? Sales, increase in the share of wallet among existing customers, acquisition of new customers, brand relevance, etc. And finally, they built a driver tree, a KPI driver tree, on how these key business indicators can ultimately be measured from different touchpoints.
So not every Like means I have somehow increased brand awareness here. This is also an excellent example that not every metric is a meaningful KPI. From the individual marketing and sales activities, which metrics can I generate that result in meaningful KPIs that show what marketing contributes to my overall business goals? After the conceptual work, infrastructure was built. I had different OSRAM and touchpoints, websites, social, certain apps, order systems, paid channels, and organic media. And from all these channels, so to speak, pull out the relevant metrics to fill the driver tree I mentioned. In principle, we built a technical infrastructure, an ETL process, and ultimately the data. In this case, it was a Microsoft architecture, the data stored in the database, specific metrics formed there, and then in the last step dashboards made available. These were then used both at the central division organization and in the countries for the operational management of the digital marketing and sales activities. Aggregated dashboards and reports built up for the administration, up to C level. In principle, they could then look at a higher aggregation level in real-time and see what contribution marketing and digital sales ultimately make to achieving the higher-level business goals. This was then rolled out worldwide, countries were trained, and a genuine business asset was eventually created. Of course, the whole thing can also be used to raise awareness of the relevance of digitization in the respective business model, for example, through command centres that are hung in the reception area or prominent places. This is an excellent example of how a field of action, so to speak, was identified from the strategy. Ultimately, all marketing and sales employees at headquarters and in the national companies were successively guided to the topic of data-driven business management via dashboards. Jan, you have a second example based on the strategy.
Jan Stöckel: Yes, thank you very much. Yes, we are also very proud at Osram that we are hanging in the foyer of the control centre. So anyone who has the opportunity to enter will see the dashboard that Albert has just seen hanging directly in the lobby. Now to another example, the company is a medium-sized medical device manufacturer active in supply, post and operative, and preventive. Mainly compression products, but also bandages and prostheses. And here, we have been supporting the company for years in digitalization, digital strategy and conception, and consulting e-commerce are the main topics. And when you are active in e-commerce worldwide, there are also structures and, let's say, conflicts with certain retailers. And it often happens that retailers don't adhere to the sales specifications and, for example, sell products outside their country or through channels that are not intended. To monitor such situations and view the price structures worldwide, we have designed and set up an international price monitoring system for me, which shows the individual departments in the company, from C-level to sales staff, which products are offered in which channels at which price. Of course, the media are mainly digital because offline channels are difficult to monitor in this area. But even in the online space, we have identified various areas.
It starts with marketplaces just as it does with large retailer structures. In the case of marketplaces, we monitor Amazon, for example, and for retailers, it's a topic like Zalando, where we look at what prices are. In addition to medical products, Medi also has a fashion and a sports line, also in focus. That means we monitor six product ranges. And we also look at competitor products. How do they compare with medi products in terms of price, and how are prices developing there as well. In other words, retailers are monitored in particular, and retailers who do not adhere to certain specifications in terms of channels are placed on specific blacklists, for example. And they look at whether the retailers are still active or whether there are products that cannot be sold online. The legal situation, for example, is such that in some countries, prescription products are not allowed to appear in regular trade, such as in Canada, and it is awful if the products then appear. This could result in the withdrawal of prescribing rights, and that is, of course, what we want to prevent. And accordingly, the monitoring offers the parties involved an excellent overview so that they can intervene early and prevent worse.
As you can see, the whole system is also cloud-based with Microsoft technology. As Albert said, we work together with a partner who crawls everything, and we do the real voodoo. That means we aggregate the data, enrich it, do currency conversions, and add additional product information from medi to the data to categorize or classify. The dealer data is expanded, and assigned groups of people are identified. For instance, we also know in the dashboards which sales areas belong to which employee and the like. And can then play things out via various dashboards and reports. We essentially work with Power BI here, i.e. the playout system. But we also create PDF reports for particular areas where there is no Power BI license. This started as a relatively lean project and has expanded into a great dashboard and project. And medi uses that daily to see derivations for the business.
As a second example, I have brought another IoT project today. IoT projects are usually data projects. Because what do machines generate? Data, of course. Here is the example of the Multivac company, a manufacturer of devices, producing packaging machines in particular. The unique feature they make the trade, i.e. food trays that consist of two films. A thicker lower film is formed via heat processes, and a transparent, often translucent film is placed on top with an imprint and closes the product. And the challenge for multivac was to ensure that the movie fit together and stuck together perfectly. And you, as the customer, are not left standing at the end, as everyone knows with cheese packaging, and can't get it open, but instead that it runs together ideally in the best case. The machines have always been equipped with sensors, and we are now recording the data. We have analyzed how this can be optimized in a project and a digital strategy. We have also built an interface for our colleagues to make specific settings. And we are constantly analyzing via sensors how to work with which temperatures and pressures so the packaging is closed ideally, of course, because it is usually filled with gas. This means that the food must be genuinely and permanently sealed for shelf life. But how can the product then also be opened relatively easily by the customer? Various technologies are used here. If you are interested, please contact us, and we will put you in direct contact with our IoT colleagues. But as you can see, data has also basically been an important area to do the project successfully. I would now hand it back over to Albert, who will give us a little overview of Sky.
Albert Brenner: Yes, the topic was data-driven marketing and sales at Sky. And the question was, how can we at Sky make even better use of the data for very different fields of application? Ultimately, it's about winning new customers, so to speak, i.e. targeting. It's about optimally serving existing customers and showing the possibilities that the respective Sky product offers for the customer. And ultimately to make suggestions, so to speak, as to what other options exist. The challenge is, especially if you are already relatively advanced in data-driven marketing and sales, how do I create additional leverage? And ultimately, the question here was how can we use data management systems, DMP, customer data platforms and the like to optimize the data-driven use cases that we already operate today, so to speak, and increase the leverage on the one hand. And on the other hand, to serve and build up additional use cases.
Whether it's the example of medi from Jan, where I can simply optimize my market management and my risk management, or an instance at Sky, where I can optimize particular acquisition activities, in other words, always show the direct impact on the business in the data topics. That is one of the recommendations there. Basically, how do you proceed in a data-driven excellence approach? From our point of view, there are six modules.
One is a description of the current status. Where do I stand concerning data? What data is available where, what is the data quality, what is the consistency, how usable is it? Functional in the sense of technical usability and usable in legal usability. To build up a data map in principle. And then, in code, to see what optimization possibilities there are? That is, I may have my data, which is already of good quality, but which would allow more information to be obtained if I were to use data from other sources, so-called second and third parties. So second-party data, data from cooperation partners. Third-party data, data that I acquire, buy-in. It can be geographic, temperature, weather, or whatever, which ultimately makes your data treasure even more valuable. Where is this data located, in what technological infrastructure are we located? And how do we manage data today? Who is responsible for data management, quality optimization, master data management? This is an important story for many companies or an important question. What does data governance look like? Is everyone allowed to use any data, whenever? So the keyword is contact management, also controlling that different departments do not permanently address the customer. I would say, monitoring data in the sense of what are we allowed to do with this data, for what have we received the opt-in for this data? So that you are on the right side. The second big step is to consider the data we have, or perhaps also generate through second and third-party data. What can we do with the data? And of course, there is a top-down approach in the sense that the data strategy should support specific business goals. That means, where can we help with data, where can we start with data? But also the other way around, to take a look at the data and which use cases can we ultimately generate from it? Then prioritize the use cases. There are many things you can do. What is the most innovative way to start?
The question is, what is the impact on my business goals? But also, what effort? Let's start with big wins. That means agreeing on a good use case portfolio developing a concept for how I ultimately use this data. One aspect that is always relevant from our point of view is the topic of touchpoint tracking. That is, how do I get data from all customer contact points - marketing and sales, as we saw earlier - that is where there is the most significant value lever, so to speak, for the topic of intelligent data. And here, of course, the contact points to the customers play a decisive role. That means building up cases like the one we saw earlier at OSRAM. How do I now manage to bring together the data from the different topics? Keyword customer data platform. At Sky, a data management platform came into being or came to fruition. We optimized the customer journey and the channels at the upper level in the final step. I think we've already seen a few good examples of this. And then, in the last point, of course, the question of how do I manage this whole issue? There is a business side. Who is responsible for it? How do I set up the structures? How do I build up the capabilities?
And on the other hand, there is also the technical side. What is the role of IT, perhaps also of external hosting providers? And, in principle, to create a structure that keeps the topic of data management in the running process. Jan, perhaps very briefly, also against the background of the time, what would be a typical procedure?
Jan Stöckel: Exactly. We have already touched on the typical procedure in some places. Some topics have already been discussed or presented in the projects. We usually start with the strategic issues at the customer's site, looking at use cases. Or, as Albert said, we first derive what use cases would make sense to implement from the company's goals. And we collect these in a backlog, then, of course, also discuss based on the leverage effect, mainly marketing sales topics, in the final impact service strategies and the like. This is highly dependent on the company's business model. Not everyone has such an end customer contact and has a customer journey as a topic, but also partly other issues that the company is concerned with. Once we have identified several use cases, we look at the company's maturity level, as Albert already mentioned. Where does the company stand? Has it already collected enough data to conduct a concrete analysis in the use case and make deductions that can be automated? Or do things still need to be built up here regarding organization and systems, resources and competencies, and technological requirements? And then, the use cases are examined again; what can be implemented relatively quickly, where do conditions have to be met? And we then derive a roadmap, an intelligent data roadmap. In other words, we try to use competent evaluations to identify assets relatively quickly or to identify topics where a company can benefit relatively quickly from data analysis and also celebrate initial successes.
We usually rely on MVPs, i.e., minimum viable products, lean projects that are as small as possible, that we suspend, and that I would say are implemented in 100 days. To also lead the organization internally, to show that it is possible to implement use cases with data that also have success and direct impact on the business. And b, to constantly develop them further and push them forward in short iteration cycles in an agile manner. And of course, this requires a specific architecture. After the roadmap concept, we look at what requirements, as mentioned before, what preferences does the company have? Is it cloud-based? Are there partnerships with big companies like Microsoft or Adobe? And there are a variety of products that we then identify and also like to implement. This means that we have a relatively broad development team of almost 400 developers at diva-e who are familiar with different systems and are happy to help with implementation. And then, the data use cases are implemented as an MVP, which is ultimately made to run in the company. This means that it must be determined who the product owner is, i.e. who is responsible for further development, communicating this throughout the company, establishing which areas work with the system, which delivery data, which consuming data, and what effects this has on the process and what must be taken into account. And then, of course, the colleagues who work with the system are trained. There are small training briefings, if necessary, by establishing new dashboard solutions such as Power BI, such as Click & Co. Of course, the company stakeholders must also be given training on how to use these analyses and what kind of derivations can be made. And finally, if the whole thing has been tested and is successful in certain areas, nothing stands in the way of an international rollout. That is purely from the project implementation side. If you take a look at the technology, how does it work? It's essentially the same. As already mentioned, we have to look at the data in the majority. Is it in silos, or does it need to be consolidated and harmonized? And then, we look at what possibilities there are, for example, for consumption. Some companies provide direct interfaces, such as adversity, Fivetran, or a segment. These enable the consumption of data from different sources. But you have to know that certain providers like Facebook and Microsoft, and Google, if you tap into Adverts, for example, regularly change their interfaces, you would have to reprogram them constantly. And such providers then offer these standard interfaces in the end. And you can simply consume them and then store them in a database on Azure, on AWS or directly at Google if you have an affinity for the cloud.
And from there, we move on to optimization and the following process, which is to think about dashboards. The aim is to develop dashboards based on the KPIs and present them so that they are ultimately readable and accessible for the end-user to understand. They can derive the information directly from the beginning and then move on to deeper analysis in the third step. And perhaps also to introduce automated processes later in the best case. In other words, this is where the topic of AI, which Albert already mentioned at the beginning, comes into play. That is, via statistical methods A or Databricks, integrated with the Azure Cloud or other systems. For example, we try to recognize similar patterns via statistical methods. This is a central topic, for instance, in the IoT projects, to then make deductions and automatically control processes, warnings and the like.
In some cases, this is not so much witchcraft, but it is also dependent on having a particular relevance of data in the first place. This means that there will also be a time lag between the dashboards and last-minute enrichments, the build-up of data until you can then really carry out automated processes. In the last area, the business is then optimized.
This means that there are not only automated slots and so on, but there are also direct rejection topics. This can go as far as optimizing Google ads in the end, for example, by using Target, a product from Adobe that allows you to address target groups. Or Bloomreach offers a similar outcome. Or you can optimize your SEA, that is, your campaigns. With intelliAd, we also provide a solution or an analysis for Amazon, for example. That's on the subject of products and technology with which we work, for example, or exemplary work. These are certainly not all that is, but there should be a few references from projects we already used. And I would now hand over to Albert, who will give you another example of how you can start quickly with us, what the next steps would be and our recommended entry into the topic of intelligent data.
Albert Brenner: Exactly. In a nutshell. What we can offer because you've seen data excellence, strategy, use cases, technologies. We've seen it in the four examples. You could make the series even longer. Conclusion: there are many starting points, possibilities, and a certain complexity. And what we would try to do with a sort of starter package, we call it Data Opportunity Workshop, which we offer you.
Three things we do there. The first is to look at which use cases are initially relevant for a company in the size, industry, customer, product, and go-to-market structure. Second, where do you stand in terms of data maturity? That's where we would do a quick assessment. This short assessment then ultimately helps to prioritize the use cases. And thirdly, we would then work with you to implement a use case directly. And what always makes sense is a dashboard that provides transparency about what data exists and how it can be displayed. Which correlations exist between data. Then you have to choose an area, is it marketing, is it sales, is it service, whatever. And then, we would build a standard dashboard, so to speak, in a kind of co-creation sprint. So a very compressed approach where we do use cases, maturity and already a use case as a dashboard development. Relatively little upfront work is required for that.
A one-day workshop and then ultimately a follow-up. I think it's a perfect, delicate, small approach that goes quickly, and for a company already, I'll say, shows the possibilities world, how you can get in the direction of data excellence. And that is ultimately our goal. And with that, I would like to thank you again very much for your participation and attention. And Angela, whoever would like to get in touch with us afterwards, so to speak, with Jan and me, here are our contact details. Angela, but we wanted to have a quick look at the survey.
Angela Meyer: Yes, exactly. The participants just answered the question of whether they already use big data. And among our participants, 50 per cent already use big data in their company. Seventeen per cent are not yet but are interested in doing so in the future. And around 33 per cent are not currently using Big Data.
Albert Brenner: There is still much convincing to be done. I can only really warmly recommend that you once again think intensively about how data can be used to provide competitive advantages as a strategic asset. And the other 17 per cent, good decision to go ahead. And what's nice is that we still have over or 50 per cent of the participants who are already active there, so to speak, implementing activities use cases. We have various analyses from our different circles, and we see that we now obviously have a very, very good maturity level of participants here in the webinar. I'm happy about that. Because in the industry average, we are still a bit below the 50 per cent. And 43 per cent of the companies have Big Data approaches, companies with more than 1000 employees. So we have a very mature target group here. I'm pleased about that. And perhaps there are already qualified questions that we will be happy to answer.
Angela Meyer: Yes, exactly. So you still have time to ask questions. I would also ask questions, even though we are already over time. We are doing a rapid round of questions, I would say. And now I would like to take up the question.
How long does a project typically take at your company?
Albert Brenner: Okay, good question. The lawyer would say it depends. The economist apparently would, too. So, it depends. Of course, it depends on where we start. If the data use cases have not yet been determined, if the situation identification is not yet available, i.e., where are which types of data, in what quality are they lying around, and also what do I want to do with the data, keyword use cases, and then ultimately to implement use cases, that takes longer, let's say. It can take as long as five or six months to implement the first qualified use case in such a project. But if we have other prerequisites, i.e., the preliminary work has already been done, Jan, what do you think we need there?
Jan Stöckel: Yes, if we have concrete requirements and data available, it is quite possible to get a project live in three months. We saw that with the media project, which was also a time horizon. With the first MVP, it is then usually pushed further. And a project like that is a continuous runner. That means that new requirements arise, which is generally also constantly expanded. Because the market is also changing, new players are joining, and it is continually being continued and expanded. As Albert said earlier, the idea is that additional data, such as weather data, can also potentially impact the area of medical products. For example, analyses have already shown that when the winter season starts with skiing and the like, the sales figures for bandages rise sharply. But I don't think we need a Big Data analysis to do such things. But something along those lines would be the next stage of expansion, for example, at media.
Angela Meyer: And now there came another question.
I wonder if you could explain the Data Majority Assessment in more detail. How does it work exactly? And approximately how long does it take? Do you do interviews or?
Albert Brenner: Yes. By the way, we would be happy to drop another short e-mail afterwards, and then we can perhaps also send you one or two slides. So, we have a very structured procedure in the maturity model very briefly. We do two things. On the one hand, we do interviews. That is correct. And we also look at data systems, and we look at data sets. We have a wide variety of analysis tools, so to speak, with which we then access the data sets that can be made available. And to establish initial analyses and ultimately form an overall picture from them. So interviews alone are not enough. I want to look at the quality of the available data. For that, I have to look at something like that. But you can do something like that in a compressed form in six to eight weeks.
Angela Meyer: That's right. So all other questions and details by mail or directly by phone. You're welcome to contact them directly, and they're open to any questions. Here's a quick note about our other webinars that we host weekly around SEO, e-commerce, and content. Feel free to keep signing up for our webinars.
We look forward to your participation. And now, thank you both. Thank you, Jan and Albert, for your insights, and thank you to the attendees for your time, and say until next time.
Albert Brenner: Thank you.
Jan Stöckel: Thank you, until next time. Ciao.