The costs of acquiring new customers are getting higher and higher due to ever-increasing competition; for this reason, it is of vital importance to maintain the ability to retain customers, to develop their potential with constant and personalized communication, and to optimize marketing activities costs. Marketing automation fits into this context, owing to the growing opportunities offered by artificial intelligence, stands as a tool that enables management of marketing activities in an integrated manner: a real strategic asset for companies who want to be competitive in the market.
To analyze the advantages (framing it within a business context in which it can be a lever for results generation), it is important to understand the potential through a clear definition and an overview of critical capabilities, with a special focus on trends of the coming years, thanks to a growing introduction of Artificial Intelligence as a support tool to analysis (augmented analytics) and automation (RPA).
What is marketing automation
Marketing automation is the set of activities and technologies that allow you to automate and optimize the management of marketing campaigns across multiple channels. We can see marketing automation as a tool in support of marketing teams to manage activities in complex contexts where target audiences and touchpoints are not limited to a handful of combinations. Marketing automation is also extremely useful for allowing those involved in marketing to focus their actions on strategic activities with high added value, even in “simpler” and less complex contexts.
Critical capabilities of marketing automation
Among the critical capabilities (or rather, those necessarily present functions) of marketing automation, we find:
- Lead Management, a capability that allows you to manage the entire database of contacts, analyzing its composition and working, in a more or less automated way, on multiple operations on it;
- Lead & User Tagging and Segmentation, which allows you to segment contacts and users’ databases, both manually and on the basis of behaviors observed on the analyzed touchpoints;
- Lead & User Scoring, which allows you to associate and calculate scores based on lead data and interaction data (website navigation, social and e-mail interactions) aimed at qualifying each individual contact, not only in terms of segmentation and interests, but also of propensity to purchase or contact;
- E-mail & Touchpoint Automation, which allows you to define contact workflows of individual users on various touchpoints (e-mails, social networks, chatbots, etc.) based on the behavior of each user and on the values of tags, segments and scores. Typically, this capability is expressed through visual systems for the construction of journeys or workflows, and by WYSIWYG (What-you-see-is-what-you-get) editors for the construction of messages;
- Reporting, which allows you to extract status reports from the system (availability of contacts, distribution for a specific analysis dimension) or performance reports, crossing the available metrics and dimensions. The systems also typically enable to program customized and differentiated reports for the various stakeholders based on templates;
- CRM Integration, which allows integration with the main CRM systems through bi-directional synchronization, in near real-time.
Completing the list, some functions that are now becoming a de-facto standard of marketing automation:
- PPC Campaign Management, that is the set of functions that allow you to manage simple PPC campaigns on the main paid channels (Google Ads, Facebook Ads, etc.), ensuring correct tracking through the insertion of UTM parameters in the campaign, and the integration of campaigns cost data in integrated and full-funnel reporting;
- Web analytics, or rather the set of functions that allow you to analyze user behavior on the company’s digital properties, crossing this behavior with the classic metrics and dimensions of marketing automation, in order to develop complete analysis of the performance of the campaign. The most advanced platforms also integrate more or less sophisticated attribution models that take into account all the touchpoints used;
- Website personalization, developed through client-side scripts that allow the personalization of the user experience, based on previous interactions managed by the other marketing automation functions (trivial example: custom home page with message in line with the one received by e-mail that has been sent based on a workflow);
- Landing page management, which allows you to create landing pages using visual and drag-and-drop editors starting from templates already designed for lead generation purposes, without the need of technical skills.
Marketing automation of tomorrow, by the use of AI
Until a few years ago, marketing automation was almost always associated with the sole function of workflow automation and contact flows design via e-mail. In reality, the space of marketing automation is much wider, and the constraint to expansion of automation in other fields has been primarily of a technological nature. Many of the repetitive marketing activities often have inhomogeneities which have made them difficult to automate through a set of conditions that are easy to define or simple string comparison criteria, or even through a predetermined definition of the parameters with which to set contact flows or campaigns launches. Thanks to the progressive democratization process of AI, caused by an ever more important diffusion of algorithms, libraries and machine learning and deep learning technologies applied to marketing, today marketing automation is taking up a significant role. Due to the evolution of Natural Language Processing (NLP) technologies, marketing automation can, for example, work effectively in the classification of lead behavior and automatic tagging without a manual definition of the labels, but by recirculating users back to specific interests thanks to the text analysis of the pages they have visited. Always through NLP technologies (and in particular text generation and summarization technologies), marketing automation can generate paid advertising campaigns and develop small customizations in communication, starting from data gathered on the web, on the basis of which to shape your behavior. Analytical capabilities are also enhanced, thanks to an integration of analytics and AI (the trend that Gartner collects under the name of “Augmented Analytics“) which helps marketers to be able to carry out analysis that were previously the prerogative of data scientists, identifying anomalies in automation flows and campaigns, and describing particular patterns in data through advanced conversational analytics systems that allow to distill insights and suggested actions (prescriptive analytics), on the basis of user behavior data. On the other hand, the SaaS systems typically used by companies, including marketing automation software, are becoming more cooperative with other systems through API interfaces and synchronization functions with in-cloud Data Warehouse. This allows them to work on the advanced integration of these systems, thanks to no-code and low-code automation tools ( example: Zapier) or even through the increasingly popular no-code AI tools (such as Levity). In this way, marketing people can have full freedom in managing their data and activating them on different channels, focusing on strategic planning, without limits of imagination on automation. However, artificial intelligence, applied to marketing activities automation, has a peculiarity in comparison with the classic approach with defined workflows and manual optimization: it needs data from which to learn. Well-structured data, in which the business goal and possible factors to consider are clear. For this reason, it becomes extremely important to plan a tracking and measurement strategy that is consistent with the marketing strategy. With this perspective, measuring will no longer “just” become crucial for analysis, but also to make things work correctly and allow campaigns to perform.
Introducing marketing automation in the company
The implementation of marketing automation within a company does not involve an ordinary path as it might seem, and it often clashes with some underlying problems that companies should face before starting an automation journey. This type of consideration is valid for automation paths of any kind, but especially in case of marketing automation, since the effect on customers, and therefore on the perception of the company and on revenue becomes immediate. We can outline three fundamental prerequisites for successfully implementing marketing automation within the company, without risking it to be just an additional layer with little business impact – boiling down to an advanced way to manage e-mails:
- high digital maturity;
- strong integration of marketing and sales;
- well defined processes – even better if well formulated.
In many cases, especially in companies that are not natively digital, it is difficult for the whole company to fully satisfy these prerequisites. For this reason, it can make sense to identify a section of the company (a business unit, a practice, a product or a work team) that lends itself to quickly introduce automation and use it as a case study for testing and optimization, then moving on to the extension of the application of the model to the whole company, only at a later time. In four steps, we can define a journey to the introduction of marketing automation, ranging from goal-setting to its implementation in the company through an iterative cycle of setup – use – analysis and improvement.
The first and important step to successfully introduce marketing automation in the company is undoubtedly the path to identify, set and communicate all goals. In this phase, it is important to involve the various stakeholders in the working group, who will be affected by marketing automation, in some way: from those who will read data, to those who will work on the automation paths, up to the client-facing team (sales, account, customer success) that could optimize their work or obtain useful insights thanks to the implementation of automations. In this phase, it will be important to collect the various requests and define a mapping of the short, medium and long-term goals. The latter must be quantifiable and measurable, so that they can be verified. In addition to the targets, it will be important to define users’ needs with respect to the automation path – in this regard, the user story tool can be effective to achieve the purpose. Through the user stories, we are able to write functional specifications putting the user at the center of our design path, by using a form like the following: As a [user description], I want [feature or action] So that [goal or value for the user]. The goal-setting process has the function of directing subsequent actions and enabling a high-level involvement by the entire working group; in this phase, clear and effective communication becomes crucial, so that the introduction of processes and technology are not likely to become a fake add-on in the company, but a pervasive and transformative element. We need to get out of this phase with strong (and documented) clarity on:
- measurable short, medium and long-term goals;
- specific needs for the implementation of these goals, preferably using user story tool.
Technology stack analysis
Another key step is the as-is analysis, which identifies all the already adopted technologies to store and analyze customer and user data, and manage the relationship with them. In this phase, we must focus both on proprietary technologies that are the heart of the business (e.g. in case of an eCommerce, we could analyze the state of the art on the technology that manages the website), and on support technologies and tools (e-mail client, any CRM or customer data storage systems, etc.). For each tool it is important to analyze:
- level of use by internal users;
- existing data – and any redundancies or asymmetries with other systems within the company;
- methods of communication and data exchange with the system (are there any APIs or databases which you can interface with?)
In this phase, it may be important to be followed by a martech partner, such as ByTek, in mapping the existing systems, in order to evaluate the possible involvement of system integration figures or companies, or the use of no-code or low-code integration solutions.
The choice of technology
With clear goals in mind and a broad visibility of what already exists, we can proceed to define the needs in terms of functions (lead scoring, tagging, etc.) and system characteristics (how much it must be modular and/or customizable, if it must be open source, on premises or in the cloud, etc.), paying particular attention to the necessary integrations; in some cases, the multitude of already existing systems within the company can guide the choice towards a solution just because it is easy to integrate. We also clearly define any price and implementation time constraints and, at this point, with all these elements in hand, we can start scouting for solutions, by characterizing them in relation to the following aspects:
- system features;
- integrability and interoperability;
- estimated implementation times.
Now we can proceed with the actual choice of technology. This is one of the key steps: technology is chosen almost at the end of the design path, not at its beginning, in order to adapt the choice to the needs and not vice-versa. In that case, it could generate useless legacy, implementation problems and poor system adoption.
Introduction of the solution in iterative mode
At this point, let’s try to understand how to incrementally embed the solution in the company. We can start from a part of the process (e.g. only cold lead management) or a part of the company (a division), or use both dimensions. For the choice, we can use the PIE framework which helps us to prioritize which part of the company and process to “automate” first. The PIE framework plans to assign three scores to each choice:
- Potential, that is a measure of the improvement that can be made;
- Importance, or rather a measure of the importance and impact of the area within marketing (for example, if product A measures 70% of the company’s turnover, then it will have higher scores than product B)
- Ease, which is a measure of the simplicity of the system implementation on that area.
We can then assign a score to each part of the process on each part of the company (for example, contact workflow automation of demo requests from the website for product A, automation of ads campaign management for product B), but keeping in mind the different perspectives when we assign scores: people, processes, technology. The process is extremely iterative: we start with an activity planning, then we implement it, we measure the results, we scale up this planning, and then, once the potential improvement has been achieved, we proceed to another area. If the company has a very strong structure, it might make sense to be followed by a partner such as Talent Garden, which assists companies in corporate transformation, by innovating organizational models through the design methodology.
The process of introducing automation can change the way in which the company approaches marketing and sales, if done properly. Through artificial intelligence, the possibilities will multiply, and companies – both the “digital native” ones (already well ahead), and large corporations and SMEs that will have to evolve – will increasingly require such tools and processes, in order to be competitive. For this reason, there will be a much greater need for consultants, professionals and marketing specialists with the right mindset and skills to face this challenge. This is the reason why in ByTek and Talent Garden we are investing resources in the research, study and construction of pioneering technological and training products that can assist future companies and professionals in this exciting journey.