How not to drain Marketing investments: Voximplant case study

The guys came to me with a clear request: to combine data from different sources and develop clear and visual reports on finance and marketing on them. We started working on the project in May last year. Now Voximplant continues to improve its data, and my analysts are refining dashboards in accordance with these changes.

I talked to two representatives of the customer about how the work with my team was developing and what results it brought:

● Ivan Alekseev — Head of Voximplant Analytics Department. He supervised the project for the development of marketing dashboards;

● Yulia Grosheva is the financial director of the company. Under her leadership my analysts made financial reports.

Company data is stored in different accounting systems: in advertising cabinets, in CRM, in the financial system and corporate database. It was clear that the internal opportunities of Salesforce CRM are not enough to build full-fledged marketing reports — data from multiple sources is needed for end-to-end analytics.

For example, to see the results of all step of the works with leads it should be clear that:

● how much money is spent on each campaign
● what sources do leads come from
● how are they handled by managers
● how many of them become real clients
● how much money they bring to the company

The answer to the first and second questions is in different advertising cabinets. On the third and fourth can be found in CRM, on the fifth you need to look in the financial system.

Of course, the company knew how much money they spent on each advertising campaign. And it even knew how much money the advertising campaign received for them. But why so much? How to increase this amount? To answer these questions, we need end-to-end analytics that will combine all the information in one tool. Voximplant took up this task without waiting for possible losses due to errors in the distribution of the marketing budget.

Data from different accounting systems was also required for full financial analytics. But the main difficulty here was in the method of calculating metrics. According to Financial Director Yulia, it is not enough to know which metrics need to be tracked. The main thing is to determine the exact rules for their calculation. I couldn't test different approaches on my own and do it quickly.

It would seem that since the company uses Salesforce CRM, it was logical to choose Tableau as BI platforms. But this idea was abandoned almost immediately and Microsoft Power BI was preferred to others:

● it's easier to "connect" it to all accounting systems - not only CRM
● some employees already had experience working with the product
● the market for Power BI support specialists is more suitable

The choice of the Institute of Business Intelligence as a contractor was also not thought about for a long time. There are not so many integrators on the Russian market, so the choice, according to Ivan, in principle was small. “I didn't want to go to large companies — it's both long and expensive. In addition, there is a serious gap between those who sell the BI system and those who then implement it”.

In addition, it turned out that both project managers from the customer's side had already encountered me before. Ivan knew me from his last job (I provided the dashboards master class for them), and Yulia took my online course.

By the way, Julia is worth saying separately. From her we received not only wishes for the financial dashboard, but also the MVP of the product. This is, in principle, a rare case when the client himself makes a "trial version" of the dashboard. An even rarer case is that the CFO of the company does it himself.

We did not spend weeks coordinating estimates and waiting for a technical task from the client. We followed the Time & Material scheme: conducted interviews, identified needs, discussed the rate for hourly work and estimated the approximate amount of work. We immediately agreed to develop a pilot version of the financial dashboard "based on" the MVP from Julia.

In parallel, we also worked on a set of dashboards for marketing. To do this, we conducted several interviews with the stakeholder in English. Our experience in digital marketing, the skill of working with metrics and the ability to offer calculation options without going into technical details came in handy here.

Many people in the Voximplant team know both English and marketing metrics. But when one thing overlaps on the other, difficulties arise that require time to figure out the details. We need specialists who will "get" the necessary information from the stakeholder, coordinate solutions and offer the result in the form of a technical product.

Sometimes when the client sees the dashboard for the first time, he starts to panic: "That's wrong here, your Power BI thinks wrong!". We returned to the original data and it became clear: the question is in their quality. This has never happened with Voximplant: the team understood that it was necessary to work on the data. Therefore, in parallel, I cleaned and configured data in CRM and other accounting systems. This has already helped to get results that were not there before.

So the company detected cases when the client was not stuck to the manager or was attached, for example, to a former employee. This finding has improved both analytics on counterparties and work with them: now managers receive information about such situations immediately.

The company has been actively using dashboards for only a few months, but it has already managed to improve the quality of data in its systems and has received a base for more complete analytics.

As Ivan says, one of the results of our project and parallel work on the quality of the source data can be considered the "development of stakeholders". They treated data from storage systems with distrust. The appearance of dashboards forced us to take a fresh look at the data in principle, to realize their value. Now, if a stakeholder has a question, he looks at the reports in Power BI. And he trusts this data 100%.

Marketing dashboards in Voximplant are currently used by about 20-30 people. These are all those who work on lead generation: event managers, smm specialists, curators of webinars, targetologists, marketers. All of them watch reports in Power BI constantly and regularly. They hold weekly meetings, discuss the data received, draw conclusions, adjust their work and make decisions.

BI analytics helped to set up and optimize the lead generation process. Employees saw, for example, that the percentage of high-quality leads is very different by region. We took this into account, adjusted advertising campaigns in the regions, changed the focus, and now they are getting more leads that are converted into real customers.

Step by step stakeholders and top managers are getting used to the regular exploitation of financial dashboards, and clear metrics and visual indicators contribute to their work.

Here is an example from Julia. The insight was the division of clients into groups with different amounts of revenue. Here's what it helped to understand:

● which clients bring a small part of the revenue;
● what is the difference in Churn Rate between customer categories;
● what is the difference in LTV between customer categories?

Such analytics is the basis for making new decisions. It is already possible to understand why the indicators are exactly like this, and think about how to influence their improvement.

The metaphor about the product and the cuisinator is not mine, its author is the CFO of Voximplant. It was this comparison that came to Julia's mind during the discussion of the results of our project. You take the product, she says, run it into a food processor and quickly get excellent semi-finished products. Fortunately, like other employees of this company, Julia understands at the same time that the quality of the original product is on the customer's side.

An interesting point: some stakeholders used to think that new tools would be needed for a "beautiful picture" on the dashboard, which means new investments. We were surprised when we saw that this was not the case. And we noted the work of designers who are able not just to work with Power BI, but to place accents so that the necessary information attracts attention and is remembered.

Voximplant became an unusual customer for us. Both in terms of interaction with us, and in terms of the level of culture of working with data. It is rare to say that about anyone: yes, this is a company with a Data-Driven approach. It understands the value of data and knows how to use it for business growth.