Prakash Kolhe, Founder of Cymetrix, walks through how AI is transforming digital marketing across every stage of the funnel. From personalisation at scale and predictive analytics to real-world architecture using Salesforce, this is the full transcript of the webinar.
Introduction
We recently hosted a live webinar on one of the most talked-about topics in business today: Artificial Intelligence in Digital Marketing. The session was led by Prakash Kohli, Co-founder and CEO of Cymetrix Software, who has over 30 years of experience in CRM, marketing technology, and enterprise software.
This blog covers the full webinar. Everything Prakash walked through, from the opportunities and challenges AI presents in digital marketing to each stage of the funnel, the real-world architecture, and the questions the audience asked at the end.
🎥 Prefer to watch instead of read? Catch the full webinar recording on YouTube: AI in Digital Marketing: Watch Now
AI in Digital Marketing: No Longer Optional
As the title suggests, we are talking about artificial intelligence in digital marketing. In today's world, AI is being used across every sphere of life, and it is evolving every single day. We are always trying to take a snapshot of where it stands. And especially in the field of digital marketing, it is no longer an option. It has become a necessity.
The question is: how can it help across the different areas of digital marketing, what are the tools and platforms available, and how do organisations actually enable themselves to use AI in their marketing journey? That is what this session is about.
Today, people like you and me, we expect brands to almost read our minds. We want faster responses, relevant recommendations, and experiences that feel personal at every step. But on the other hand, for businesses and marketers, it creates a real challenge: how do you deliver that level of personalisation and still improve conversions across the funnel?
Let us start with the strengths AI brings, the challenges it has posed, and the opportunities it has created for professionals in digital marketing. AI is no longer optional. It is a competitive necessity. If you don't use it, your competition will be using it and will definitely benefit from it. You may be a loser in this game. So you have to study, understand, and start using artificial intelligence in your digital marketing journey.
Organisations that balance technology, ethics, and human creativity will gain the greatest advantage. Many organisations and individuals fear that AI will either take their jobs or eliminate creativity. But actually, human creativity is the most important ingredient in digital marketing. AI tools are to be leveraged for that creativity. It has taken the whole thing to a different level, and of course, you have to use it very ethically, because it could be a boon as well as a curse. If you use it in the right way, it will be a boon.
The opportunities AI creates are significant. A lot of personalisation is now possible. Before AI, we depended entirely on the skill of the digital marketer to personalise, and personalisation at a mass level was simply not possible. Mostly, it was done at the segment level, which is a very high-level cohort of users and may not address all the microcohorts within it. But using AI, that micro-level personalisation has now become possible. A lot of work automation has become possible. Creating campaigns, running campaigns, and managing different campaign journeys, which were earlier done manually, can now be automated. It is available 24x7. Efficiency has increased. And the insight into what is happening with the customer has grown enormously because of AI.
Of course, it has also posed certain challenges. There is a privacy risk. Most of the time, team members resort to open-source LLMs that are freely available and then post organisation data to do research or content optimisation. That definitely poses a risk of putting your organisation's data out in the open. There is a cost involved. There are skill gaps. Teams may not have the right skills to leverage these tools and platforms, so there is a need to upskill and bridge that gap. And there is a lack of transparency and uncertainty about what is going to happen.
The key trends picking up because of AI are hyperpersonalisation: personalisation has gone to the next level, now possible at the micro segment level and even at the individual level. Chatbots have almost replaced all web interactions. Most customers, especially on e-commerce websites, can now use bots to interact and find what they need, with very little human interaction required. Predictive analytics has become possible. If the data is there, a lot of predictive behaviour on customer trends and results can be derived using AI. And of course, the new trend is ethical AI, which means using AI in the right way so you don't go down the wrong path.
The Digital Marketing Funnel: Stage by Stage
In digital marketing, we have been using the funnel traditionally. It has different stages. First is awareness, where we reach out to and attract potential customers through email marketing, website SEO, and other methods, making our potential customers aware of our offerings. Then comes consideration, where potential customers get attracted, ask for additional information, and explore further. Then you convert them and close that customer. After that begins the cycle of retention, cross-selling, and advocacy through referrals.
It is no longer just "digital marketing" as a single word. It has been divided into the different stages of the customer journey, and based on each stage of the funnel, the approach to digital marketing has to be different. Because of different tools available and because of artificial intelligence, there is far more leverage of technology possible at each stage. Let us go through each one.
Awareness: Attracting the Right Audience
In the awareness stage, the customer is on a journey of discovery. They are discovering what their needs are and starting to explore. From the digital marketing point of view, the focus has to be on attracting and targeted outreach. You have to understand who the targeted audience is and reach them. That is how this stage of the funnel becomes successful.
Now, how does AI help here? One area is the execution and optimisation of digital ad campaigns across multiple channels. The number of channels through which digital ad campaigns need to be done is increasing day by day: social platforms, your website, and now even different LLMs across which you may have to place your digital ads. The execution and optimisation, because of AI, can be taken to the next level through automation and having the right targeted audience identified. It works 24x7, and no team member is required to attend to it. Once you do the configuration in the right way, it runs automatically. This reduces the manual workload for marketers significantly.
On budget optimisation specifically, you always have a limited budget, though it results in good leads. But how do you optimise that budget? Traditionally, those decisions were taken by the marketing team. Now, because of AI, there are tools where this budget optimisation is decided by the tool itself, depending on the results, what CPA (cost per acquisition) you want to limit, and how different ads can be bid across different platforms. All that automatic decision-making and placement can now be done through AI. You can even set a ROAS (Return on Ad Spend), and the tool automatically takes care of placing ads within that and delivering results. Then there are tools for designing campaigns for specific goals like online sales, lead generation, driving store visits, or optimising CPA and ROAS.
AI-powered content generation is another big area: video, image, audio, all of it can be generated using AI. Video editing, like backdrop remix or inpainting, is possible through AI platforms. Lip syncing is possible. 3D texture generation is possible. Plagiarism detection, which is especially important from a legal standpoint for larger organisations, can be done automatically using AI. Drip campaign automation can be automated and intelligently placed based on results. Built-in SEO optimisation can be done, and keywords can be automatically added. Brand-wise customisation is done. Especially for product-related videos, brand customisation can be maintained across different ads. And multilingual support is possible. You don't need to record your videos or messages in different languages. Through AI, it can be done automatically.
Personalised outreach through micro audience targeting is very important. AI identifies highly specific audience groups. It can identify across different visitors to your websites or campaigns. It can even find lookalikes. If there is an anonymous visitor, it can figure out who that person could be. New prospects can be identified that resemble your best existing customers.
Creative optimisation can be done automatically by testing different variations. You can find out what kind of email header, body, and message works, or even for chatbots, you can test different knowledge inputs and see how the response changes for a particular micro audience. Over time, the AI platform learns and executes better, which definitely results in more conversions at the awareness stage.
One of the most important capabilities is identifying and analysing the emotion, language, and content performance across channels based on data, something that humans have clear limitations with. Earlier, advertising messages were not created from an emotional understanding point of view, and that emotional understanding is a very important aspect if you want to create the right first impact on your target audience. You need to understand, based on what segment a customer is coming from, what their emotional needs are, and what language should be used in the content.
At a broad level, there are five types of emotions: pride, anticipation, fear, joy, and trust. For example, if somebody is coming with a wedding anniversary context, they may be approaching with a joy-type emotion. Within that joy, you can break it down to micro emotions: excitement, fascination, gratification. Accordingly, you can position your ad based on how the person is approaching, what they are chatting about, what email they are sending, and what content they are clicking. And this is possible only through AI; manually, it is not possible to analyse at this micro level. Every word you use in a message can address a particular emotion, and that can be carefully drafted through AI tools.
The key tools used at the awareness stage include Google Performance Max and Meta Advantage Plus, which create automated audience targeting, creative optimisation, and AI-driven ad placement with automated bidding. Jasper AI and OpenAI are content generation tools that you can decide what kind of content you want, what kind of headline for which segment, what body, what message, for blog posts or social content at scale. Persado is a very evolved AI tool for emotional language understanding. It understands the emotional state of the customer based on chat or messages and positions content accordingly. Synthesia and Runway ML handle AI video generation. Albert AI handles autonomous digital marketing and cross-channel ad placement.
Consideration: Engaging and Nurturing
In the consideration stage, the customer is doing research and evaluation. They may have visited different product websites and are in evaluation mode. As a digital marketer, the focus should be to engage that customer once they visit your website or content, and nurture them with relevant information that is very important to make the consideration stage successful.
AI helps in several ways here. It improves sales efficiency by capturing and analysing buyer intent signals and predictive modelling. You must understand the buyer intent of the customer visiting your content, and analyzing that buyer intent has become possible because of AI. Through predictive modelling, you can even predict what the customer is going to do next.
AI can identify high-intent accounts and forecast which accounts are closest to purchasing, helping prioritise sales efforts. Since you sometimes have limited sales power available, you need to prioritise which accounts you should focus on, where is the possibility of conversion high? Those can be identified through AI, and the sales team can put in further efforts to convert them.
AI can analyse buying signals from across the web, identifying anonymous traffic and tracking keywords, content consumption, and competitor research. It can identify new target accounts that resemble your best existing customers, even when they are anonymous. It can generate conversational and personalised email campaigns to automate lead qualification because when you get too much interest, you never know if it is a serious lead or not. Lead qualification was traditionally done by call or email, but because of AI, more personalised emails can be sent for lead qualification automatically. No manual intervention is required, and it automatically tests whether this lead is serious.
Programmatic display and LinkedIn ad campaigns for targeted accounts can be automated and optimised. AI can identify which topics and keywords are driving actual revenue opportunities, you can change or add keywords based on that, and test in real time which keywords are generating more opportunities and which micro segments are getting attracted to which keywords. All of that is being fed back into the database to be used for ongoing campaigns.
Intent intelligence can be built based on conversational chat as chat is being done, AI can analyse it and find out the intent, what the client wants to do. Dynamic web content can be used to modify website copy, images, and calls to action in real time. Based on the customer and their micro segment, the website copy will be automatically adjusted, and the content shown to that user will be different from what is shown to other users or different segments. Images and calls to action can be changed in real time to drive action. Dynamic landing page generation can also be done based on customer profile and micro-segment.
The impact of personalised experiences on pipeline generation and revenue can be measured. Everything can be tracked if this personalisation were done for this particular micro segment. How did it impact the pipeline? That keeps on getting improved as the AI platform learns. It can automatically direct more visitor traffic to better-performing page variations. If a particular web page is trending, any person coming to your website can be dynamically redirected to that page. Content experience adapts in real time based on visitor behaviour, intent, and engagement signals, delivering the next best asset without manual curation, whether that is an email or a video.
GEO Generative Engine Optimisation automatically generates metadata for websites and LLMs based on learning from visitor behaviour and intent engagement.
The key tools here include 6Sense for intent data, account identification for account-based marketing, and predictive lead scoring. Drift and Intercom for conversational AI sitting on chatbots, reading intent based on chat. Mutiny and Intellimize for website personalisation and dynamic landing pages. PathFactory for content intelligence and personalised content journeys based on engagement and intent. Clearbit and ZoomInfo for data enrichment, whatever limited login details you have, these platforms fill out the full customer 360 data, and give you more insight into the customer profile.
Conversion: Closing the Deal
The next stage of the funnel is conversion, which is the most important because you have to close the deal here. You have done a lot of effort, you have brought the target audience to your website, you have made them consider your product, and now you are at the stage where it has to be converted. The purchase intent is very clear. The customer has decided they want to buy. From a digital marketing point of view, you have to do persuasion, which is the typical sales cycle, and close with urgency.
AI can help a lot at this stage. It can analyse customer interactions, all the calls done so far with that customer, the emails exchanged, and any meetings done using AI to provide actionable insights for the salesperson. It is kind of a co-pilot for the salesperson, giving a lot of insights on that particular customer. Forecasting and coaching for the sales team on what this particular customer wants, what you should position, and what you should pitch is a great help for any salesperson to convert.
It analyses team activity to predict deal outcomes, identify risk, and monitor deal health. If there are calls being made on the phone, AI can monitor what kind of calls are being made, what kind of pitching has been done, and whether it is the right approach or not. Even in the CRM, it can track what kind of visits were done, how many meetings, and how many times meetings happened. All of this data is analysed by AI to predict what kind of outcome the deal will have and whether there is any risk, maybe because of the wrong salesperson being positioned. That kind of risk can also be identified and highlighted.
It measures how much the sales representative talks versus listens, that kind of pattern on the salesperson, whether they talk more and listen less, that risk is highlighted, and coaching is done on it. It tracks mentions of competitors, pricing, or specific pitches and analyzes customer sentiment. When you need to convert the customer in real time online, it automatically suggests that this is the competitor this person has been visiting, at this pricing, they can get converted, or this particular feature, if pitched or added to your offering, will get the customer converted. That kind of suggestion is done through AI.
Speaker identification, if the customer is on a call, automatically identifies the person. If there is some history in the past, if any conversion happened, if any purchases were made, just using the speaker's voice, it can identify the customer. It can automatically capture and sync sales activity, email, calendar, and meetings. Tools like Salesforce CRM can track all sales activity and generate intelligence from that related to deal conversion.
Key tools at the conversion stage include Salesforce Agentforce, which is the AI product from Salesforce built on top of CRM, it can do predictive scoring, saying this lead has this score and can be converted, what next-step action can be done, and it can do deal prioritisation. Gong is a conversational intelligence tool for all sales call analysis to be done, as well as side-by-side coaching as well. Highspot is an AI content recommendation and sales enablement tool. If you want to convert next, this is the content that can be sent, or it will automatically send it. People.ai is a revenue intelligence and pipeline optimisation tool that maximises conversion and gives you higher returns and ROI. Dynamic Yield and Optimizely handle real-time personalisation and checkout optimisation.
Retention: Keeping Customers Coming Back
Once you have converted, the next stage is retention. By retaining the customer, you don't have to go through the whole cycle again. You also have a lot more data available, you know what the customer's different likings and needs are, and you can use that for cross-selling other products.
The customer journey here is post-purchase; the customer is already inside your organisation. Digital marketing has to focus on delighting that customer and upselling post-purchase, which also leads to good feedback and better referrals.
AI helps by doing customer engagement automatically based on intelligence. Post-service automated services, delighting customers, monitoring the actions of internet users who interact with your application, what they are doing, their intent, and engagement, all monitored and recorded. Based on that, you can judge whether the customer is happy or not, or what the issues are. The delay of seconds between insight and action is the difference between a retained customer and a missed opportunity. That timely action is very, very important.
Connect with customers across different touchpoints throughout the customer journey and real-time messaging. This is possible only through AI, because using human manpower, these things are very difficult to achieve. Personalised nudges at optimal times. If you know a particular customer's birthday or marriage anniversary date, you can gently nudge them at the right stage for additional purchases or gifts. Those kinds of nudges are possible and automated.
Even something like one of our customers had a need around vacation homes. If the customer is visiting at special events at their vacation homes, even that data is captured what services the customer uses when they visit. Next time, when that particular date comes, you could send a message like this particular spa or service is available at a discounted rate. That kind of cross-selling of your other products becomes possible.
Predictive churn is used to identify and re-engage inactive users. If you find a particular user is inactive, there is a high possibility of dropout or losing them to competition. To prevent churn, you can re-engage them by offering certain discounts or other methods. AI agents for instant autonomous resolution of inquiries, that time has gone where customers log a query, and some agent says they will respond later. Through AI agents, instant gratification can be provided wherever possible. And of course, it is a co-pilot for agent assistance; it can assist agents who are servicing the client, telling them what this customer needs, what their likes are, and what their past is. It works like a full co-pilot.
Key tools here include Braze for AI-powered engagement, Amplitude and Mixpanel for complete behaviour analytics, which can tell you what features the customer has been using, what they are not using, and how many times they are logging in. Gainsight for customer success AI helps with post-sale health monitoring and renewal prediction.
Advocacy: Turning Customers into Referrers
The next stage is advocacy. Once you have a retained client, that customer needs to refer to other customers, and that is the business most organisations are eyeing. Referral business is the best business available because your cost of acquisition is very low. It is very important that you encourage your customers to refer.
The customer at this stage is looking for loyalty and expansion. Digital marketing has to focus on encouraging and enabling referrals. AI can boost customer engagement through gamification, targeted activities, and rewards. As we know, many airlines have gamification and targeted activities referral programmes, personalised member journeys, referral tracking, and referral management. Prompting customers, partners, and employees to share, comment, and like, you keep prompting them for feedback.
As a user, if they are giving comments and feedback, we know that even today, with something like Zomato, you look at the feedback and rating given by other customers. That is very helpful in selling your services or products. Get more reviews, grow your reference pool. In-app guidance is automated in-app guidance. If your customer refers to another customer, that referred customer should be able to get onboarded automatically.
Automatically suggest features to track a visual heatmap of user interaction, showing what features they are using more, keeping an inventory of that, and using it for improving the product for that particular feature. Automatically identify and alert teams to opportunities for retention or immediate service recovery. Identify precise emerging trends from unstructured data that is available. Analyse the data to determine the underlying reasons behind customer behaviour. You can even go to the why of why a customer is clicking a particular action or is not being active. Those kinds of whys can also be answered through the data accumulated using AI.
Architecture: How It All Comes Together
Let us look at the architecture for how an organisation can implement AI in digital marketing overall.
This is one architecture we have just worked on for one of our customers. This customer is in the automotive business. In their automotive business, they had an e-commerce kind of website also available on mobile, and they wanted to design dynamic content. The content management system CMS has all those assets that need to be automatically used for updating these websites, and it also has data coming from different systems through the integration layer to bring data from other systems as well.
Google Analytics or Adobe Analytics was used for analyzing website traffic, tag management, and all of that data. Now it is important that you build complete customer intelligence, the customer 360, which is done on the Data Cloud product of Salesforce. This has been built by bringing data through integration from Salesforce Service Cloud, ERP, and Sales Cloud, which is the CRM, and that golden record is created here.
That golden record on the customer is used for marketing purposes through Salesforce Marketing Cloud, which is the marketing automation tool. This helps in customer journey orchestration, different channel integrations related to communication, campaign execution, and segmentation, keeping segments live and interacting. This is the complete digital marketing AI tool setup.
Now this is how we use the AI tools for marketing across the different funnel stages: awareness, consideration, conversion, retention, and advocacy. At all stages, there will be a generative AI foundation like OpenAI, ChatGPT, or Gemini for GenAI, and there will be AI-powered chatbots used across the website as well as other social media and interaction channels. These tools are used depending upon the stage, integrated along with Salesforce Marketing Cloud, which is the marketing automation tool. It uses data from the customer 360 in Data Cloud to be used in the campaign. At the same time, whatever campaign intelligence is generated is stored back again in the Data Cloud for the next stage, and it keeps building customer intelligence inside your CDP or Data Cloud.
Key Takeaways from the Q&A
After the session, the audience asked some great questions. Here are all of them:
Audience - Is it better to integrate AI into your existing CRM and marketing platform or use separate AI tools?
Prakash Kolhe - These AI tools can independently work, but effectively, if you want to use them, they have to be integrated into your existing CRM. And you need to have a marketing platform. Some of these tools themselves work like a marketing platform, or if you want to use multiple tools, then you can use one marketing platform and integrate different AI tools into it.
Audience - Before implementing AI marketing, what level of data maturity is realistically required?
Prakash Kolhe - Data related to the customer is very, very important. If you don't have a CDP already, that is a platform you should have if you want to leverage AI in the most effective way. These AI tools generate a lot of data and need a lot of data as well. If you want to start on this journey, whatever customer data you can consolidate, have that, and then use it. That is the best start, and going forward, it automatically starts building on that. Having a CDP is the most suggested platform for going through this journey of AI for marketing.
Audience - Within the data ecosystem, which is the most important kind of data?
Prakash Kolhe - AI in marketing is generally used for selling something. To sell something, you need the customer 360: what their intent is, what their driving factors are. Any customer data you have in the system is most important to have. Amongst all the data, the data in the CRM is the most important to start with because it has been gathered over a period of time. Then there are tools like ZoomInfo, which can enrich this data further. Apollo or ZoomInfo to enrich the data. Unless you have more insight into the customer, you cannot have more enrichment of customer campaigns and messages.
Audience - How should small or mid-size businesses approach AI in digital marketing without large budgets or data teams?
Prakash Kolhe - At the very minimum, you need a CDP where your customer data is, a CRM which has your sales cycle data related to the customer, a marketing tool where journeys are managed, and on top of that, different AI marketing platforms depending on the stage you are using. Now, if it is small or medium, this is an expensive stack if you want to implement all of it because the Salesforce stack itself is expensive, and some of these AI platforms are also expensive.
But if you really want effectiveness. Let us say you want to generate awareness-related campaigns. You need a CRM. That is a must, to have complete customer data. There are multiple CRMs available depending on your budget. Then, depending on your need, you can use any data lake or data warehouse. There are multiple options available. For marketing automation, there are multiple platforms available: HubSpot, or even Mailchimp for email marketing automation, which is more economical. But if you want to do across multiple channels like social media, mobile, and website. Then you need a tool that covers multiple channels and platforms.
You can start with just one channel. Maybe email marketing, see the results, and then scale. You can use tools like Apollo on a need or access basis to enrich your data. Unless you have more insight into the customer, you cannot have more enrichment of campaigns. The key point is: to expect outcomes, you have to spend. It needs a budget. But start small, start with one channel, build the intelligence, and grow from there.
Audience - How should marketing leaders measure ROI from AI-driven initiatives?
Prakash Kolhe - In AI, results may not be immediately available the way you might see with social media spending, where you can directly track CPC and conversions. AI platform ROI maybe requires patience. It is over a period of time. One reason is that for training the AI on your data, your offering, and your customer data, it takes some time to show results. Ultimately, you measure over the period of time how much you have spent and how much conversion has happened. Different AI tools show varying results: anywhere from 30% to 60% improvement in customer conversion is possible. That improvement in customer conversion, how much does it result in revenue, and whether it justifies the AI spend? That case-by-case calculation needs to be done.
Audience - Will AI replace performance marketers or simply transform their roles?
Prakash Kolhe - No, it will help performance marketers. It will not replace them. As said in the beginning, AI is not replacing people. Of course, it is taking away some of the manual tasks, but for creators, it is more helpful. For performance marketers specifically, if you are using different tags or keywords, AI tools can automatically suggest certain tags and keywords depending on the users. But finally, you, as a performance marketer, will have to train and control what happens. I believe it will not replace. It cannot replace. It will only help in a better way.
Audience - How can predictive AI identify high-intent users in real time?
Prakash Kolhe - To use that, you will have to use tools built for finding high-intent customers. As we have seen in the conversion stage, tools like Gong, for example. If you are using chat, this tool can automatically predict what the customer is chatting about, what their intent is, and what the prediction is for getting them converted. Tools like People.ai can help you identify which customers are high intent and prioritise them. Even suggest what content can be placed for that kind of conversion. The tool does this based on its own internal logic: historical data, different aspects of the customer, their sentiment, the language they are using, and, of course, the past history of conversions.
Audience - What about data governance and accuracy when multiple platforms are enriching and scoring the same accounts?
Prakash Kolhe - Data governance has to be there. At the back end, elements like governance and security, those have to be there. Data quality for the data coming into the Data Cloud or CDP needs to have data governance in place, not violating regulatory requirements, and ensuring it is a golden record. Those efforts and spending related to data quality are most important if you want to implement AI effectively. If you don't have clean, governed data going in, you will not get reliable results coming out.
🎥 Access the complete discussion as it unfolds here in the video: AI in Digital Marketing
About Cymetrix
Cymetrix is a Salesforce consulting partner and technology firm specialising in CRM, data, and AI-driven digital transformation. With over 155 certified consultants, delivery centres in India, and client offices in the USA, UK, Poland, and Japan, the firm works with mid-market and enterprise clients across BFSI, manufacturing, retail, healthcare, and professional services. Covering the full Salesforce platform: Sales Cloud, Service Cloud, Marketing Cloud, Data Cloud, Agentforce, and more. Cymetrix brings over two decades of Salesforce delivery experience to every engagement.
Our Salesforce Marketing Cloud consultants help organisations implement and optimise Marketing Cloud for customer journey orchestration, multi-channel campaign execution, audience segmentation, and AI-driven personalisation, turning it into a revenue driver across every stage of the funnel.
Book a Complementary 1:1 Strategy Session now
- Log in to post comments
