Martech Architecture

According to observations, the number of Martech investment enterprises "have everything but still lack" or "everything but none of them operate effectively" accounts for a fairly large part".

‍

The main reason is that there is no comprehensive view of the overall architecture, which leads to disjointed solutions and overlapping functions. The following is a more specific example with a classic case:

‍

• If you search for an email solution, you will immediately find hundreds of email solutions. You decide to "buy" quickly.
• The following month, an automated messaging solution arises. You also decide to "buy".
• The next month, you find it necessary to connect the data between the email and messaging software to avoid sending the same content to customers on two channels. So do you have to buy new software or try to use it manually with the two previously purchased software?

‍

‍

You see, only two solutions have entered such a dilemma. If you invest in a more complex and larger technology solution, things will be much more troublesome, right?

‍

The solution to this problem is scientific: when you think of an "email" solution, you have to understand that it belongs to the "Interactive Layer" and study it further. Next, you notice that the engagement layer with the concept of "Multi-channel Marketing Hub" (MMH) is a group of functions that includes sending emails and sending messages.

‍

You'll realize, "Well, it's a set of features, so you should buy the whole set, or part of it first, but make sure the first part of the purchase must be attached to the later purchase." From there, your thinking about architecture will take you to the next level.

‍

So what is an interactive layer? What is Omnichannel Interaction? Where does it fit in the overall architecture of Martech? Don't worry, in the next section I will explain Martech architecture and CRM right here. Martech is essentially a combination of technology and data to support Digital Marketing. Therefore, Martech must be placed in the context of the data.

‍

The miniature architecture of Martech can be illustrated consisting of two floors as follows:

‍

Interaction and Experience Floor

‍

The interaction and experience layer will take on the role of directly interacting with customers through touchpoints such as email, chat, website, app, phone on both online and offline.

‍

The data layer will take on the role of collecting customer information from the interaction and experience layer, combined with customer information from business databases such as ERP, POS, DMS, eCommerce, etc. The information generated by the interaction and data layer is the input to the Data layer, on the contrary, the information after being aggregated and processed by the Data layer will be the input of the interaction and experience layer.

‍

If you observe, you will realize that the data layer is the foundation of "targeting" and the interaction layer is also the foundation of "instant marketing" in the journey of sawing down CRM.

‍

When we go to combine the Sales and Service module into Martech, we will have the following architecture:

‍

Martech's expansive two-story architecture

‍

This is also the basic architecture of a CRM system. However, each industry (banking, retail, manufacturing, distribution, services, etc.) will have its own characteristics, each business also has different current conditions and orientations, so the Martech architecture or CRM for each specific business will also have great differences.

‍

In the Experience and Interaction layer as well as the Data layer, each color cell inside is a function in the Martech picture. A few common terms and concepts to help you visualize it better.

‍

• Marketing Orchestration
• Omnichennel Engagement
• Social listening và Engagement
• Unknown Visitors
• 360-degree customers
• NBO/NBA Recommendations
• Personalize the experience

‍

Marketing Orchestration

‍

The function plays the role of a conductor in managing and coordinating Digital Marketing-related components. This module manages and synthesizes digital content (forms, images, audio, video), customer data for marketing, building campaigns, defining workflows between related personnel, and even managing marketing budgets.

‍

Omnichannel Engagement or Campaign Management: is a function that plays a role in planning and coordinating interactions with customers with the goal of sending the right message at the right time on the most suitable interaction channel. The evolution of customer engagement will make it easier for you to visualize:

‍

• Send messages to customers on a single channel such as Email, SMS, etc.
• Send the same message to customers across multiple channels but there is no connection between these channels (multi-channel)
• Send messages to multiple channels but have an omnichannel connection. For example, only send an SMS when the customer hasn't read the email after three days.

‍

Omnichannel Engagement can be considered as the place to manage multi-channel journeys or campaigns (Campaign Management), including defining a flexible interaction scenario in terms of content, channels as well as when to interact to bring a personalized experience to each customer. Example:

‍

• The journey to convert customers who abandon the cart but have not paid (abandon cart)
• A campaign to invite customers to a new product launch event.

‍

Social Listening and Engagement: is a function that plays the role of listening to information and managing interaction on social networks. Concrete:

‍

• Listen to and monitor social media users' feedback on product quality, service, customer care, brand, or related media incidents. This includes listening to and monitoring businesses and competitors in the market or the entire industry in the online environment in general. This is seen as a measure of the effectiveness of a media or marketing campaign.
• Prepare content, schedule releases, and post on multiple channels and accounts simultaneously, saving time and avoiding errors. On the other hand, it is also possible to interact directly and centrally when there are comments on social networks.

‍

‍

Unknown visitors

‍

Visitors who have not logged in or have not left identification information (name, phone number, email, etc.). When you first visit a website or e-commerce page of a certain business X, even though you have never logged in or left any contact information, all your behaviors such as where to click, what products to view, what news to read, for how long... will all be saved.

‍

There is no magic here. An unidentified customer is nothing more than not being identified (name, phone number, email, etc.) according to the understanding of the physical society, but in the digital environment, when you visit a certain website, you have been "digitally identified" by cookies, IP or device identifiers.

‍

"Knowing and understanding" the behavior of an anonymous visitor has a major impact on the ability of this visitor to convert into a customer. Until you log in or leave your information, any history from previous unidentified visits will be reconnected.

‍

360-degree customer

‍

A place to show customer information in a "complete" way. Of course, "complete" here is also very difficult to quantify. At its most complete, a customer's information can be expected to include four demographic, geographical, behavioral, and psychological factors:

‍

360-degree customer information: Geography and demographics are two traditional elements of marketing and are increasingly showing little value because these values change little while shopping needs change very quickly. Behavioral and psychological factors become valuable and quality but can only be obtained with the application of technology. So companies try very hard to invest to have this information.

‍

Did you notice that the picture above looks a lot like the scope of a gun? The more information about the customer, the more accurate the sighting.

In another interpretation, 360 degrees includes customer information before and after identification. However, because it has not been standardized and defined as 90, 180 or 360 degrees, many service providers are quite arbitrary when determining that they have 360-degree customer information even though they only have some basic information.

‍

‍

NBA/NBO Recommendations

‍

Structurally, NBA (Next Best Action) will include three main categories: NBO (Next Best Offer), NBS (Next Best Service) and NBI (Next Best Information). I find that the translation into Vietnamese here is not very good, I hope you can understand the original word in English to feel the full meaning.

‍

The NBA has its own KPIs such as for user manuals, upselling/cross-selling, customer retention, and enhancing the experience. Depending on which KPI you want to push, choose the right plan. For example, after logging into an eCommerce page, searching for a product, adding some products to your cart but not paying for them, you decide to turn off your browser.

‍

A good solution that personalizes the web experience will capture this moment and allow administrators to come up with a number of options:

‍

• Display a 5% coupon for an unpaid cart order if you don't close the browser, return to the cart, and check out within 15 minutes. This is the NBO;
• Display a survey to survey customers about the checkout experience and what requirements to improve quality? This is the NBS;
• Display a message asking the customer what is the average monthly income? This is the NBI.

‍

Types of NBA

‍

There are two ways to do the NBA, either using rules-based laws, or using Artificial Intelligence/Machine Learning. I have some better explanations of Rules-based, AI/ML in the CRM Puzzle section at the end of this book, if you're interested, learn more.

‍

‍

Personalize the experience

‍

Flexible customization of information content, interface, interaction channels according to the habits and behaviors of each visitor/customer. Depending on the context, it will determine the "what" is customized.

‍

• The most common example is on the website. With the support of technology, the deployment is extremely simple and fast with a few pieces of code attached to the existing website. Depending on who the visitor is (no identification is required), the website can customize the interface (colors, images), the order of arranging buttons, the order of displaying products, the popup time of notifications, promotions, etc. in the blink of an eye (less than 1/3 second).
• Another example is personalizing the experience on email. Good email sending solutions can personalize both the content of the email, the time it is delivered to the recipient's inbox, the frequency of sending, etc. based on interaction behavior on the email channel (whether the email was opened, when it was opened, whether the link was clicked, whether the attachment was downloaded...).

‍

Personalizing the experience is one of the mainstream trends that guide the technologies of Digital Marketing. To give you a better idea, I'll take a few specific examples with two groups: Failure Experiences and Success Experiences

‍

Here's a group of examples of Customer Experience failures:

‍

• Goal-seeking advertising - Retargeting. For example, you go to Shopee to buy a refrigerator, later you go to Facebook and you will see an advertisement for Shopee's Refrigerator. The next day, you go to vnexpress and see that advertisement again. However, when you buy a refrigerator, ship it home, three days later on Facebook, you still see Shopee's advertisement → fail o You work as IT for bank A, you go to Vietnamworks to fill in profile information. Every month you receive an email inviting you to Apply for the position of M&E Worker of an electric vehicle factory → fail

• You are using bank C's credit card, and you occasionally receive an email from bank C inviting you to open a credit card → fail - Here is a group of successful Customer Experience examples:

• I will take an example in the field of Banking. Imagine that you are a customer who has a payment account but has not used a credit card of bank B, has an income of 10 million/month and travels frequently.

• One day, when you enter the cafef.vn page, you see an Ads Banner of Bank B with the content "You have been selected (pre-approval) to become a customer using a credit card with a limit of 20 million. Click here to participate." → Success if: Bank B targets you - the existing customer to offer pre-approval with an exact limit of twice the income. Conversely, an ad fails if it appears to everyone, or if the quota doesn't match.

• You click on Ads and are taken to a landing page. Oddly enough, the landing page looks like it's configured for you with a large banner with a Travel theme, masculine website colors, and a menu that's configured for easy access to credit card information. → Success if: a female colleague of yours also goes to the same landing page, but will see the hero banner about shopping, the pink color is gentle, the menu is also different... Even the interface changes in terms of the day (morning/afternoon/evening), whether you are in Hanoi or Saigon, how many degrees is the outdoor temperature...)

• When you fill in the pre-approval information for your credit card, the system knows all your personal information, so you don't have to fill it out. You enter information about finances, hobbies, habits... According to a flexible process, the system recommends the right type of card and announces the results immediately (is your application approved, how much is your level, are there any promotions...) → Success if: the survey process is extremely flexible, the next question depends on the previous question,  handle complex logic instantly without 25 waiting times from the back-office. This case applies to commitments such as reviewing documents for one minute is the standard.

• Continuing with the above step, you enter complete information but do not click the submit button but close the browser (just like the case where you shop online, put the goods in the shopping cart but do not make a purchase). 30 minutes later, you receive an email with a reminder that you forgot to click submit. If you do nothing, three days later you get a message with a reminder. Seven days later, you receive an email from the bank, but this time there is an attractive incentive program if you open the card right away... → Success if: all content and interaction channels of the bank are based on your behavior (digital body language): ignore emails, read emails, click on links, fill out forms without submitting, go to websites... are tracked and automated in a personalized way for yourself.

‍

To be able to generalize about the ability to "personalize the experience", I divide it into four levels as follows:

‍

Levels in Digital Marketing - These levels are a combination of two axes

"Channel Coordination" axis: divided into four levels:

‍

• Single: only one interactive channel

• Multi-channel: there are multiple channels but only one channel per interaction stream

• Cross: there are many channels, each interaction flow can use multiple channels but is not tightly cohesive o Omni: there are many channels, each interaction flow has a close coordination of multiple channels and focuses on a 1:1 experience with customers

‍

"Interactive content" axis: also divided into four levels:

‍

o Mass: customers are not grouped according to any criteria o Segment: each content that interacts with customers corresponds to groups by geography, gender, demographics (with families, small children, etc.) or combined with dynamic factors of time such as:  birthdays, company establishment dates, special events...

o Personalization (1:1): interactive content is personalized according to each customer's behavior and habits on a few single channels.

Cognitive 1:1: Interactive content is personalized at the highest level with information about customer behavior, interactions, and habits across all channels to create a flawless experience with a combination of artificial intelligence.

‍

‍

The combination of these two axes forms four levels of experiential anthropomorphism:

Level 1: no personalization on one or several channels

Level 2: simple personalization based on pre-given information or conditions that combine customer profile information with behavioral attributes to personalize content

Level 3: personalize content and recommendations (NBO, NBS, NBI, etc.) based on single-source data across multiple channels

Level 4: personalize content and recommendations (NBO, NBS, NBI...) based on aggregated data from multiple sources and multi-channel interaction in real time If you have a good grasp of architecture and concepts in Martech, I have good news for you. That's when you're ready to learn one of the hottest tech concepts available today: CDP. Find out now and let's go!

‍