How we did the most massive internal survey in the bank
About a year ago, we, DIT VTB, conducted a successful user satisfaction survey , covering 20,000 people - all VTB employees using IT services. This was followed by the accession of VTB24 to VTB - a combination of different cultures, habits and years of experience. To understand how the accession was perceived, it was necessary to conduct another survey - now for 40 thousand people. What we did at the final stage of the merger of banks. How we approached the survey and solved the attendant difficulties, read this post.
For a start, it was necessary to decide in what form to conduct a survey. In March, our colleagues from another division of the bank had already conducted similar activity on an outside resource, with a lot of promotion in the office. They hung banners, distributed leaflets, hung a banner on the site. As a result, only 3000 participants were able to gather. Probably, the problem was that it took half an hour to answer all the questions.
When preparing a new survey, we also thought about its large-scale promotion. As an option - on the placement of campaign videos. After all assessments, it became clear that making a good video is a long time and in our case is unprofitable. Although, in principle, I liked the roller theme, we gained experience, contacts and decided to use this format in other activities. But this is a topic of another post.
The survey, which we wrote about last time, was conducted before the merger of VTB24 Bank. From it, we took the first question so that the results could be compared directly: “Evaluate the quality of the work of the Department of Information Technologies”. And in the second question, they have already focused attention on the unification period: “Evaluate the quality of the work of the Information Technology Department at the time of joining VTB24”. There are five possible answers: excellent, good, normal, bad, and very bad. Optionally, you could leave a comment.
So that 40 thousand people from two previously different companies could quietly participate in the survey, we created a new resource with the possibility of pass-through authorization. They published the news on the internal portal of the bank, attracted employees through the mailing list, and also placed the flyers in the canteens - the places with the most crowded people.
The most effective was the mailing list, but the news on the portal also added a lot of audience. Leaflets didn’t add as many participants as they created the general atmosphere. It is logical - it’s problematic to go over a link from a paper leaflet, but when an employee sees a letter or news on the portal, he immediately understands what this is about.
Voted as a result of more than six thousand people. According to calculations, the answer to the questions took no more than 30 seconds, and with a comment - no more than three minutes. It turned out more productive than in the case of a 30-minute survey, which passed three thousand people.
More than a thousand users also found time to comment on satisfaction with various aspects of IT work - they described their impressions, emotions, and left wishes for development. From the comments, we, for example, determined that users experiencing problems with VDI, was 8 times less than last time - technical improvements have not gone unnoticed. Last time Tolik was a “sweetheart”, but now Marina, Andrey Valerievich and Sergey Nikolaevich have joined him.
The total data from Sharepoint was minimized to Excel. To create a separate system for the sake of a one-time task did not. During the first survey, we already did this, but then we evaluated it in terms of the time spent on setting the TOR to specialists for developing and running data analysis scripts. And the total labor costs are not justified. In addition, our task is half analytical, half creative.
Data for analysis were taken from different sources:
An interesting task was to determine the difference in assessments of employees of ex-VTB24 and VTB. The systems did not have such a clear sign; therefore, they were divided by indirect attributes, for example, by the mail domain. Such a division was not included in the survey results, since within the bank no one divides employees into ex-VTB24 and VTB.
A comparison was made with the results of a similar survey for the past period for all sections, since often the fluctuations are much more indicative than the final figure. All this was considered in a breakdown by business area and by region. User evaluation tied to his requests for technical support, in order to understand on the basis of which services the general opinion on the work of the DIT was formed. At the same time, we looked at what estimates the user gave on the basis of their applications. Often, the execution of all applications he was satisfied, but in the survey put a low rating. This is an important issue for a separate communication.
Thanks to the survey, we obtained a lot of data, between which we can find interesting dependencies that do not lie on the surface. This will help us develop IT services in the future.
For a start, it was necessary to decide in what form to conduct a survey. In March, our colleagues from another division of the bank had already conducted similar activity on an outside resource, with a lot of promotion in the office. They hung banners, distributed leaflets, hung a banner on the site. As a result, only 3000 participants were able to gather. Probably, the problem was that it took half an hour to answer all the questions.
When preparing a new survey, we also thought about its large-scale promotion. As an option - on the placement of campaign videos. After all assessments, it became clear that making a good video is a long time and in our case is unprofitable. Although, in principle, I liked the roller theme, we gained experience, contacts and decided to use this format in other activities. But this is a topic of another post.
The survey, which we wrote about last time, was conducted before the merger of VTB24 Bank. From it, we took the first question so that the results could be compared directly: “Evaluate the quality of the work of the Department of Information Technologies”. And in the second question, they have already focused attention on the unification period: “Evaluate the quality of the work of the Information Technology Department at the time of joining VTB24”. There are five possible answers: excellent, good, normal, bad, and very bad. Optionally, you could leave a comment.
So that 40 thousand people from two previously different companies could quietly participate in the survey, we created a new resource with the possibility of pass-through authorization. They published the news on the internal portal of the bank, attracted employees through the mailing list, and also placed the flyers in the canteens - the places with the most crowded people.
The most effective was the mailing list, but the news on the portal also added a lot of audience. Leaflets didn’t add as many participants as they created the general atmosphere. It is logical - it’s problematic to go over a link from a paper leaflet, but when an employee sees a letter or news on the portal, he immediately understands what this is about.
Evaluate the results
Voted as a result of more than six thousand people. According to calculations, the answer to the questions took no more than 30 seconds, and with a comment - no more than three minutes. It turned out more productive than in the case of a 30-minute survey, which passed three thousand people.
More than a thousand users also found time to comment on satisfaction with various aspects of IT work - they described their impressions, emotions, and left wishes for development. From the comments, we, for example, determined that users experiencing problems with VDI, was 8 times less than last time - technical improvements have not gone unnoticed. Last time Tolik was a “sweetheart”, but now Marina, Andrey Valerievich and Sergey Nikolaevich have joined him.
The total data from Sharepoint was minimized to Excel. To create a separate system for the sake of a one-time task did not. During the first survey, we already did this, but then we evaluated it in terms of the time spent on setting the TOR to specialists for developing and running data analysis scripts. And the total labor costs are not justified. In addition, our task is half analytical, half creative.
Data for analysis were taken from different sources:
- The survey itself;
- Personnel system;
- Two systems for maintaining user requests;
- Active Directory - used as an intermediate slice for re-checking and correct data binding.
An interesting task was to determine the difference in assessments of employees of ex-VTB24 and VTB. The systems did not have such a clear sign; therefore, they were divided by indirect attributes, for example, by the mail domain. Such a division was not included in the survey results, since within the bank no one divides employees into ex-VTB24 and VTB.
A comparison was made with the results of a similar survey for the past period for all sections, since often the fluctuations are much more indicative than the final figure. All this was considered in a breakdown by business area and by region. User evaluation tied to his requests for technical support, in order to understand on the basis of which services the general opinion on the work of the DIT was formed. At the same time, we looked at what estimates the user gave on the basis of their applications. Often, the execution of all applications he was satisfied, but in the survey put a low rating. This is an important issue for a separate communication.
Thanks to the survey, we obtained a lot of data, between which we can find interesting dependencies that do not lie on the surface. This will help us develop IT services in the future.