More
    HomeAnalytics AI & AutomationHarnessing the power of weblogs for customer analytics

    Harnessing the power of weblogs for customer analytics

    -

    Sponsored: Should you develop a customer analytics platform from scratch or buy a pre-built solution? The type of data you ingest could hold the key, writes Intent HQ’s Benedict Enweani.

    Your customer data is the most valuable asset you have. Use it well and you stand to greatly improve revenue, customer loyalty and margins. Yet there is a data source so rich in customer behavior, it matches the value of data held and used by over-the-top service providers like Facebook, Google and Amazon, yet most communication service providers (CSPs) continue to ignore. This data source is weblog data.
     
    Transforming this into tangible business value can be hugely challenging. With so much data to deal with – a lot of it noise that must be cut through to find usable insight – CSPs face a major challenge, but also major opportunities.
     
    Take control of your weblog data and the possibilities are huge. This could include increasing your response rate through personalized marketing campaigns, improving your data monetization efforts, cutting churn and delivering higher conversion from upsells and cross-sell offers that boost customer lifetime value. But how do you achieve these goals?
     
    At the heart of this will be your customer analytics platform. And in many cases, the solutions CSPs already have in place are not likely to be good enough to handle the scale or level of detail needed to process weblog data and deliver truly personalized campaigns.
     
    Therefore, you need more advanced technology. There are a number of ways you could go about this. You could build your open platform, look to embrace general cloud tools from one of the major suppliers, or opt for a prebuilt solution that has been designed specifically to meet the needs of weblog data and the telecoms sector.
     
    For many CSPs, building their own platform or adding to the tools they have will seem like an attractive option, as it gives them a high degree of control over the proprietary technology. However, it also comes with many issues, such as the unfamiliarity of weblog data, the time required to build it, ongoing maintenance efforts and the resources required to build a best-in-breed solution. However, these can all be avoided by buying from a specialist third-party provider.

    Managing the complexity

    The first challenge is dealing with the scale of the issue. Large CSPs will need to ingest trillions of data points into the system in order to get results, so artificial intelligence and machine learning tools will be essential in managing weblog data at this level and converting it into valuable insights.
     
    This translates directly into time. To build a complete customer analytics platform in-house is a major commitment, and it could easily be three years or more before you have a solution that’s fit for purpose. Even then, it will require constant iteration and improvement in order to keep it working effectively. This is time and energy that even the biggest CSPs may not have.
     
    You could save some time and complexity by turning to off-the-shelf tools from major cloud providers. But in this case, you’ll still have a lot of work to do to take these generalist tools and adapt them for the unique needs of weblog data and the communications sector, especially once you factor in any necessary add-ons. You will need specialist data science resources and continuity of these resources, which in the current labor market are in huge demand.
     
    Another alternative, buying a complete solution, can offer a much faster time to market. However, it is important to find a vendor with the experience and technology required to create value from weblogs. With the right partner, you could be ready to start deriving actionable insights from a single customer view that combines your weblog data with CRM data, billing data and other sources of data in months, or even weeks.
     
    With no need to build the software or build bespoke ML Operation teams yourself to get started, you could spend less time creating the tools needed to analyze the data and more time focusing on the insight itself.

    Addressing the costs

    The time and complexity of building from scratch also translate directly into additional costs, and again, these can quickly spiral out of control.
     
    If you’re building yourself, there are a range of expenses to consider, from dev costs to compliance functions. Meanwhile, turning to mainstream cloud providers may also lead to a variety of unexpected customization costs as standard solutions have to be reworked and adapted to meet the specialist requirements of weblog data and the telco sector.
     
    There are also human expenses to consider. Does your existing team have the skills and resources available to execute this strategy? If not, you’ll need to spend heavily to bring onboard the right people – something that’s not easy in a competitive job market.
     
    For example, the average salary for a data scientist or data engineer when changing jobs rose by 20% last year, reflecting how in-demand these skills are. Even if you can find the right expertise, the time taken to get them up to speed with the business may add more delays and costs to the project.
     
    What’s more, it’s not just the costs associated with building a system that need to be taken into account. There are also ongoing expenses associated with maintaining and upgrading the solution. This type of project can never be signed off as ‘complete’ – there will always be improvements and additions to be made as technology advances and customer understanding changes. World-leading CSPs such as Verizon and O2 Telefonica understand this and license this technology from Intent HQ rather than build it themselves.
     
    If you simply declare your project as finished, it will not be long before you drop behind rivals and have to invest again to catch up. However, with a bought solution that specializes in the complexities of surfacing insight and human behavior explainability from weblog data, you can be sure that all this continuous development is being handled for you.

    Making customer data work for everyone

    There’s one other factor that you may not consider until it’s too late if you’re planning a build from scratch. This is ‘who’s going to use this platform’? In order to get the most out of your customer data, people throughout the business need to be able to ask questions and interrogate the data to draw their own conclusions, whether they are in marketing, customer experience, IT or analytics teams.
     
    If you are planning an in-house build, the chances are the task will fall to your data science team. And if you do this, you’re going to end up with a product that’s designed to be used by a data scientist. This will make it harder for other users without high levels of technical knowledge to use the system, and may even result in data science teams acting as gatekeepers, with any request for analysis having to go through them as they’re the only people who understand the solution.
     
    This greatly harms both the efficiency of the process and the quality of outcomes you can expect. However, with a bought solution that has been built from the ground up to enable intuitive train of thought analysis, anyone can pick up and start exploring the data regardless of their background or role.
     
    Opting for a prebuilt solution that can handle and add value from weblogs therefore greatly streamlines the process of getting powerful, effective customer analytics programs up and running, and ensures a quick return on investment. One that’s easily configurable to your needs and delivered as a cloud-based Software-as-a-Service proposition, one with high levels of customer data privacy baked in that provides a quick and easy way to start taking advantage of customer data and boosts the overall value of your weblog assets.
     
    A solution built from scratch, on the other hand, will be expensive and time-consuming, with few guarantees of success.
     
    Find out more about how you can put weblog data to work with our Intent Platform and AI technology, and put powerful insight in the hands of CSP teams.

    The author, Benedict Enweani, is Chief Commercial Officer at Intent HQ.