Digital transformation & Robot Process Automation

There are 70 probabilities out of 100 that you have opened this newsletter using a Microsoft email application and there are still very high probabilities that this is the application you use most frequently, together with MS Word and MS Excel.

This behaviour is often influenced by the fact that we tend to use a minimum portion of these applications’ potential, as if not having to send faxes - not to mention envelopes - nor use an eraser or a calculator all seem to us in themselves satisfactory signs of an indisputable progress. And it is actually true that progress has been made, but it is still early days.

One of the cornerstones of digital transformation is the branch of information technology called Robot Process Automation, or RPA. This includes programs (BOT) which, either automatically or on demand, can perform sequences of repetitive functions in a quick and effective way 365 days a year, 24 hours a day. We will not go into the details of RPA, but we would like to give some examples of how it can increase efficiency while improving the quality of some of the most common office procedures.

Attachments

How many do you send? How many do you receive? How many do you return with amendments and comments? How many do you forward? If your answer to one of these questions is “more than two per day”, then think how it would all be simpler if attachments were stored in a shared OneDrive repository and instead of sending them as attachments and having to download them to introduce amendments, you would receive them as simple links accessible with a click in a secure manner, as intrinsically protected by viruses. In this way there would also be an end to the multiplications of versions which often generate mistakes, or at least doubts.

All this while saving time, bandwidth load and disk space. With a simple BOT it is possible to move attachments to OneDrive and send them as links without changing the way you do things, but bringing to the organisation all the relevant benefits. Multiply this by the number of people working in your organisation and consider the economic impact of such a technology.

What about sustainability? 3 emails generate the same CO2 as that produced by driving 1km by car and a server produces 1 to 5 tons of CO2 per year.

Forms

Do you ask information to your clients, define discount policies, approve expense reports? Forms are a useful tool as they allow to structure information in such a way that it can be managed by automated processes. Consider a discount request which instead of being expressed as a free text is categorised in a form with four options “10%”, “15%”, “20%” and “NO”.

You could reply with a single click which would automatically send an authorisation email amending the offer and sending the updated fee proposal to the addressee. The advantages are obvious and would further increase by integrating the analysis of information on the client (is it a bad payer?) or seller (is it in line with the targets?) in the process.

Meetings

The pandemic has added momentum to remote working and brought applications such as Teams to the fore. Now, how many steps do you need to take to organise a meeting? Or, in case an email exchange becomes obviously unproductive and a call is needed, which are the steps you need to take?

Again, actions like contacting via Teams the person who sent us the email we have open on our screen, schedule a group call for a working team sharing in advance the documents which will be discussed, record attendances and keep track of the time spent in calls are just few examples of tasks which can be automated with simple BOTs.

The partnership with Automation Anywhere

In order to create BOTs, Grant Thornton Italy established a partnership with Automation Anywhere, which realised one of the most adopted software, ranked by Gartner among the global top leaders of the industry. First of all, Automation Anywhere provides a complete software platform offering both the development environment and the runtime environment, on premise and in cloud. Moreover, it includes various components. The workplace allows to automate, innovate and transform any recurring digital activity with a BOT.

The IQ BOT allows to acquire and transform non-structured data also through AI and machine learning algorithms (it can also decode human handwriting). Finally, BOT Insight provides a data analytics solution which allows the reclassification and presentation of data.

A further interesting aspect of the Automation Anywhere platform is the Discovery BOT solution. Through this function, a specific BOT uncovers processes that can be automated, thus avoiding high consultancy expenses. The identification of processes that can be automated and the implementation of BOTs are thus sped up. It is actually possible to record users’ activities, report on business processes, support the analysis of processes to uncover automation opportunities, generate blueprints and automate them.

With a little familiarity, users will be autonomously capable of atomating their activities simply by activating functions to register their actions.

Data Driven Company

In 2006 the mathematician Clive Humby declared that “Data is the new oil”, i.e. the new most important raw material. We believe he was right, as well as those who later expanded on this concept saying that data, like oil, are quite useless if not refined and transformed in something that can be used.

Equally right were those who later clarified that refined data are useless if not included within a Data Driven Decision Making (DDDM) process. We fully support this vision and we consider Data Driven those companies in which data are collected, classified, analysed and then used to make business decisions in all business divisions.

It is actually not a technological approach, but one that requires the employment of technology, due in particular to the fact companies use different platforms, at different stages of maturity often not completely integrated: from Enterprise Resource Planning (ERP) systems used in core processes such as bookkeeping and warehouse management, to websites and social channels, to more advanced platforms such as Customer Relationship Management (CRM) and Master Data Management (MDM) systems.

Depending on the corporate systems and processes, it is necessary to define a data collection strategy and identify an aggregation platform, i.e. the data lake where raw data are collected and the data warehouse where data are structured and ready to be used for a specific purpose. A physical separation is actually not necessary, but rather a clear logical distinction in a sole repository, ideally on cloud.

The key step is the refining, which implies the identification of Key Performance Indicators (KPI) which can represent and communicate in the most effective way the corporate mission and the relevant trend.

For example, if revenues and margin are always valid as KPIs, indicating their priority has an immediate impact on the target of individuals’ activities: should we sell more or increase the margin? If we reduce the margin, will we be selling more? With the current margin, how much do we need to sell to reach the expected profits? And if we were to increase payment terms from 60 to 90 days, which would be the impact on sales and cash flow?

Data refining and data visualisation of KPI in a simple, interconnected and easy to understand way is assured by business intelligence tools. The offer is wide and - in our opinion - not strategic; the important thing is structuring dashboards in an effective way.

There are three key elements to consider:

1. Summary: KPIs need to be visible on a single screen, with trends highlighted in colour
2. Grouping: for each KPI, a screen with more details showing all related elements
3. Drill-down: upon request and for further levels, it is possible to go more in detail.

What discussed so far is the 1^st Step of the process to follow to transform a company into a Data Driven Company, the basis to start making a more intelligent and profitable use of data, i.e a predictive use of data. This means using historical data to forecast future trends and understand how to influence them.

A key role in building predictive models is played by Artificial Intelligence (AI) and Machine Learning algorithms: what is important to know is that the effectiveness of these algorithms is directly proportional to the amount of available data, which leads back to the importance of a wide-ranging data collection and of high performing data lakes.