Complexity can be reduced by designing guiding user-friendly tools. For this reason I was part of a team whose goal was to understand users, their problems, and ultimately come up with an improved solution. The case study is part of my master thesis and followed the Lean UX design methodology tailored for enterprise environment.
Platform
Web
Services provided
- Research
- UX & UI Design
Year
2016
Thesis link
View publication00.
Getting started
In order to comprehend the users and their challenges, the team (composed of two UX Designers and one Front-End developer) had to explore their environment first. The research team was not familiar with the telecommunication domain, nor with the enterprise environment. During one month and a half, the team acclimatised with the organization by interviewing local stakeholders, getting familiar with the sales strategies and products, participating in case study analysis and observing users using the indoor simulation-based tool.
The organisation
The Strategic management at the HQ defines general sales strategies and guidelines that are communicated to the execution management in the regions. There are ten regions and each of them adjusts the strategies according to its market. The adjusted strategies are then conveyed to the sales teams in each country which are part of that region. The Technical Sales Support(TSS) department receives support queries from the sales teams and comes up with solutions to solve them. The objective of the sales teams is to quickly propose solutions that are competitive from both technical and economical (cost) perspectives.
The problem
Employees at Ericsson working at the Technical Sales Support (TSS) Department operate complex simulation-based software that can be used to evaluate and compare the performance of various telecommunication products, configurations, and deployment options.
The software was created for R&D purposes and handling it requires a deep technical understanding. However, colleagues in the sales teams from different regions could benefit from the capability of the software. The tool could help them communicate solutions and estimate the costs for the customers.
This would eventually increase the volume of sales. However, from the negative quantitative (increased number of support requests) and qualitative (frustrations and complaints) feedback incoming from the sales teams, it became clear that it is difficult to use.
The simulation tool
The software renders radio propagation models inside a building, given basic building and network parameters. The original indoor simulation tool was a rudimentary UI built as an add-on on top of the simulation engine typically used by the R&D department.
A distinct challenge of the tool is that it requires processing resources for simulations, such as disk memory and time. Achieving a good optimisation depends not only on the several configuration iterations of the technical parameters, but also on the understanding of the customers and business needs, along with acquiring the necessary data. As such, a tradeoff between time, simulation accuracy and cost estimations has to be made.
Hierarchy of needs
The primary need of the sales teams from the regions is to be able to approximate the total costs of such deployments and predict performance based on the chosen configuration. To facilitate this process, they use the indoor simulation software tool. To stay competitive on the market, the secondary need of the sales teams is to make cost-efficient adjustments to solutions to reduce the costs, while still preserving a good performance. This would help the team gain competitive advantage and thus, close more deals with the customers. However, this requires the sales teams to obtain detailed data about the building where improvements have to be made and data from the network of the operator. The data can be acquired from the operators. The operators, on the other hand, are not always enthusiastic about sharing their network data, because it exposes their pain points. Additionally, it is costly to obtain specific building data, before a deal is closed. Therefore, fulfilling this need is challenging. A third need is to even further reduce the costs by employing an Ericsson design technique for indoor-outdoor network management, which minimizes the amount of network resources to deploy a solution. However, this requires a good understanding of the business case as well as the environment of the building.
The design process
As mentioned earlier, this project heavily relied on Lean UX. Lean UX is a methodology introduced by Jeff Gothelf. It applies the Lean Startup principles of Eric Ries for UX design. Instead of talking about requirements and documents, the methodology implies a multidisciplinary team that frame discussions around solving problems by delivering desirable experience for users that fit the business goals. The Lean UX methodology describes an iterative process which consists of four steps. I will describe each of the steps in the following sections.
01.
Assumptions
Once the team had a better understanding of the context, the tool and the organisation, we focused on better understanding the users and their needs. Following the Lean UX design process, we declared users and business assumptions and began creating proto-personas – “best guess of who is using the product”.
Problem statement
The TSS indoor tool was designed to help people in the region to promote and sell products. We have observed that the service is not meeting the support expectation from the regions which is causing sales go slow (not as good as expected) and creates confusion and stress amongst employees both, in the regions and HQ. How might we improve the support (guidelines and tools) so that regions can handle more questions by themselves and at the same time have fewer projects that need to be escalated to HQ?
Business and users
The problem statement was filled with assumptions and to dissect them we used the business and users assumptions worksheet template. Framing ideas on a small piece of paper in a limited amount of time and space estate set the authors to phrase and organize their thoughts in a structured way.
Prioritization matrix
By having an overview of the project in front of us, we could easily scan and identify project risks. We classified assumptions according to the Prioritization matrix template. The assumptions that involved a higher degree of obscurity and risk were given a higher priority. These assumptions are located in the top right corner in the prioritization graph.
Proto-personas
Based on employees roles and the organisational chart, the team assumed that there are six types of users. They were classified based on their type of activity (reactive VS proactive) and based on locations (Headquarters, Region, and Sales Teams). We then created a proto-persona profile for each of them.
02.
Concept & Prototype
At the same time our research team was exploring the organization and its environment, a new concept (Five-step model) for conducting sales was proposed at TSS. Given the problem statement, prioritised business and user assumptions, proto-personas and the proposed concept, we assembled hypotheses. Hypotheses are more specific descriptions of the assumptions that define the workflow for experimentation.
Five-step process
The five-step model emerged as a necessity to help the sales teams better understand the customer pain points and strategy. It extends beyond the scope of the original tool, which was focused on a predetermined building. It provides support also for other strategic decisions, such as selecting the target buildings, to ensure that the customer (operators) will get a good business case for its indoor investments. The argument behind the process was that the structured process would eventually increase the agility of sales teams to come up with an estimate in regards to customer’s demand and at the same time reduce the number of iterations required to achieve a good price/performance ratio. The model is a theoretical concept that combines several region practices. Its applicability in the field was not tested.
MVP
A Minimum Viable Product (MVP) is the smallest artefact that helps the research team test assumptions and validate hypotheses. A MVP should include only as much time, effort, and fidelity as are needed to generate useful feedback. For our initial efforts to test the Five-step model and to generate feedback I sketched on paper a sequence of screens for step no. 3 – Building prioritization, which we assumed has the highest risk based on the Prioritization Matrix. Below are some of the rough sketches.
03.
Experiment
When the MVP was ready, the next step was to test and confirm our hypotheses. Several emails were sent to targeted employees matching two personas and within a day we received replies from candidates which were interested in the experiment. We arranged 7 Skype interviews with candidates from 6 out of 10 Ericsson regions.
Interviews
The interviews were modeled to find out current sales practices and then trigger discussion around them by presenting the new sales-model. Sketches were used along with the slides to talk not only about the process, but also about the UI for the tool.
For each interview, two research members were participating; one was leading the conversation, whilst the other was taking notes. At the same time, the interviews were recorded to reference back if needed.
04.
Feedback
The interviews provided valuable feedback for the research team. Whilst we did not set exact metrics in place to evaluate whether any of the three hypothesis is confirmed or not, the notes taken during the interview suggested the confirmation of all hypotheses.
Findings & learnings
The Lean UX process helped the research team structure and test assumptions about the new proposed Five-step model. From the first iteration, the research team learned that this model suits better some regions than others. Additionally, the team learnt more about personas.
Another observation is that the Five-step model is something that would work for new engagement with the customers. From the interviews, it can be inferred that these kind of engagements are typically taking place in new emerging markets such as India or South America. On the other hand, saturated markets such as those from Europe or East Asia already have in place old indoor systems and, therefore, the new products are typically acquired to solve smaller pain points. It was also communicated by several interviews that the operators have capabilities and they know what they need to solve their problems.
On the other hand, emerging markets lack these capabilities. They often expect the companies to come up with solutions and display the benefits. Therefore, having in place a sale process, such as the one proposed, the sales team will provide valuable insights for their customers.
Another finding was that the Proactive types of personas are the right user target for the Five-process model. Practitioners are involved later on the stages of this model when the business has been set up. However, they have to be aware of the decisions taking so that they can come up with the corresponding technical solution to fit the business cases.