Artificial Intelligence in Sales

AI is already omnipresent in our personal lives. Our experiences are enriched because of its application in different areas.

Possibly without even realizing it, we use AI enabled technology in form of video recommendations, product recommendations, voice assistants, facial recognition to login and many others features in our daily lives.

AI is already saving us time, while also leading us to content and products tailored for us.

In workplaces too, AI can boost automation to save time and augment human capabilities to achieve better results.

Gartner in its post Top CRM Sales Technologies for the new realities of selling Covid-19 world has outlined 5 applications of AI in Sales processes.

• AI in Sales forecasting
• AI in Deal Scoring
• AI in Conversation Intelligence
• AI in Pricing technologies
• AI in Coaching enablement

This eBook gives CXOs and Sales leaders all they need to know about AI-guided selling and implement it successfully in their organizations.

"Business leaders believe AI is going to be fundamental in the future. In fact, 72% termed it a business advantage.” - PwC

1. What is AI-guided selling

Traditionally, we only look at activity on a specific deal, or the specific lead and make decisions on next steps.

Decisions like “what is the best document to send”, and “when is the best time to contact” are left to the salesperson’s best judgement.

We make forecasts by assigning a fixed conversion probability for each stage of the deal.

Example: New is 10%, Qualified is 20%, Ready to close is 80% etc. Similarly, we score leads by assigning fixed points for specific characteristics, or behavior.

The traditional sales process doesn’t have a systematic approach to learn from the past to improve the future results.

AI-guided selling bridges this gap.

AI is essentially just a simple process of analyzing historical data, finding patterns of causes that lead to specific outcomes, and drawing inferences that can be then used to make predictions and offer recommendations.

In addition, we now also have means to interpret sentiment automatically from phone transcripts and emails.

AI programs do this by applying Machine Learning (ML), and Natural Language Processing (NLP) capabilities.

Machines are very good at sifting through data hundreds or thousands of times in a short period of time.

AI can help answer questions like these:

• How much of an influence does the client's industry
• have on your win rate?
• How about location?
• How about the salesperson assigned to the deal?

There are many other factors - Number of engagements, Sentiment, Documents shared, How objections are handled in calls etc.

All these influences the outcome to varying degrees and AI can sift through historical data to find correlations and causations.

With sufficient historical data, AI programs can find patterns and test various hypotheses to see which factors have a higher degree of influence and provide predictions and recommendations.

Some features masked in AI uniforms do not actually use machine learning but utilize available data on the deal and contact to provide timely smart alerts (ex: System can send an alert if a deal is idle or stalled in a stage for more than 5 days).

"67% of sales reps miss their annual quota" - TAS Group

2. How does AI help a sales rep

Lead & MQL Stage

Profile Enrichment

Sales reps usually spend time on LinkedIn or other resources researching prospects. Tools that can bring in useful details given an email address or company name can be very handy and save time for sales reps.

Lead scoring

Sales reps lose a lot of time chasing poorly qualified leads. By reviewing engagement behavior and profile characteristics of leads, AI applications can identify characteristics in leads that lead to higher conversion. Engagement on emails, websites, social channels, along with profile data, are all factors that can help predict the likelihood of conversion.

Duplicate record alerts

However sophisticated your system might be, duplicate records do enter the system. Very often this is due to slight variants in the text used for the same contact or organization or deal. Programs can watch for such duplicate records and alert sales reps so that information is not scattered in different records.

"52% of companies only believed they had client data whose reliability rate was over 75%. While only 23% of companies believed that they had reliable lead data." - CSO Insights

AI can be used to improve results in each stage of the Sales process. In the pages that follow we will take a closer look at how AI can influence each stage of the sales funnel.

SQL Stage

Best time to contact

Improve the response rate by sending emails or making calls when your contacts are most likely to be willing to engage. Based on their past engagement with your team, the system can suggest suitable time-slots to call or email.

Email assistant

When responding to an email, AI can suggest documents or templates that Sales reps can insert with a click which is called Document recommendations. As the deal progresses through different stages, certain documents such as case studies, comparison documents, or ROI documents might help your prospects as they evaluate your product. AI-guided applications can look at historical data on the success rate of different documents and make timely recommendations.

Idle contact reminders

Selling is the process of building a relationship. It is important to keep engaging with your prospects, even those that do not have an immediate need. Idle contact reminders can help Sales reps keep in touch with their prospects.

"15% is the increased possibility of connecting with a prospect if you call between 8-11 am."

Deal Stage

Deal score

Knowing the likelihood of success on a deal is important to take corrective actions ahead of time, and to generate predictable forecasts.

Deal alerts & Task recommendations

Besides document recommendations, Sales reps can also benefit from other tips that overcome hurdles. For example, the system can alert:

• Sales reps if a decision-maker is not yet identified but the deal moves into the ‘Qualified’ stage.
• Managers if the client profile fit is low for a deal, and they need to step in to offer guidance to sales reps.

Action-Items from emails, calls & meetings

How often have we missed follow-up tasks that we promised? By analyzing the communication in calls and emails and using NLP, AI applications can pick out the action-items to make it easier for sales reps to follow up.

Price optimization

What is the right discount to give to a client? Should a discount be given at all? AI-based systems can suggest the right discounts or price variations based on data from past sales.

"50% of sales time is wasted on unproductive prospecting" - The B2B Lead

Customer stage

Upselling & Cross-Selling

When is it the right time to upsell? Or what other products might your client benefit from? Using AI, you can identify opportunities to expand your existing client relationships.

"85% of executives believe that AI will enable their companies to obtain or sustain a competitive advantage, but only about 20% have incorporated AI in some way, and less than 39% have an AI strategy in place" - MIT

3. How does AI help a sales leader

Deal Insights

Conversation Alerts and Deal Executive Brief

AI tools enabled with NLP can provide sales managers an executive summary on the deal, giving her a birds eye view of all the the activity on the deal, and positive and negative signals.

The tool can analyze through call transcripts and emails and look for phrases categorized by administrators (ex: Pricing, Playbook, Competitors, Features) and provide a summary.

Sales managers can see if the sentiment of the prospect is moving in a positive direction, and click through to see specific conversations.

"55% of the people making their living in sales don’t have the right skills to be successful." - Caliper Corp

Forecasts

AI-based forecast

Some salespersons are good at assigning the probability, and others are not. Some are overly optimistic, and others are cautious. Traditional forecasts often prove to be inaccurate due to these human tendencies.

Sales managers can get more reliable forecasts by applying machine learning models that consider all the engagements related to a deal that could influence its result. Don’t just go by the hunch!

As with anything related to AI, the accuracy of forecasts over time as more data becomes available to it.

"102 Days is the average lead to close length" - Salesforce

Teaching & Coaching

AI can transform Sales managers into Sales mentors

If you are managing a team of 10 sales reps, it is impossible to manually review calls even if each one spends 3 hours on the phone per day (that means you have to review 30 hours in your 9 hours even if that is the only thing you do). AI-based tools that apply NLP can provide stats such as Mood, Talk-to-listen, Longest monologues, etc.

AI-enabled tools can sift through recorded calls and emails and flag those that fall short for review by sales managers (aka coaches). Managers can see which reps need assistance and areas on which they need assistance (ex: agenda setting, discovery questions, objection handling..). When reviewing calls/emails, managers can cite examples and guidelines to speed up the learning process of the sales person.

"81% said that the accuracy of their data could be improved by capturing quality contact info from people they meet or email with." - Introhive

4. What data is essential to realize all the benefits of AI-guided selling?

Some of the AI-driven features that require historical data are:

Deal scoring

Some components of Deal Score, specifically Fit Score, Authority Score, Engagement Score are determined based on how the attributes of this deal compare with previous won/lost deals.

Task recommendations

Recommendations on what document to send or what action to do next at a specific stage of the deal will be determined based on the outcomes of doing those actions in past deals as they progressed to closure.

AI-driven forecast

By considering AI-enabled deal score, forecasts can be more done on a more objective basis.

These mentioned features require data of the past deals (won or lost) for building and testing prediction models.

"82% of B2B decision-makers think sales reps are unprepared" - Blender

Here are some of the key attributes that rely on data from the past deals:

• Deal Pipeline name
• Deal entered stages and time spent in stages
• Deal Result (won/lost)
• Deal Amount
• Owner name
• Related Org industry, size, region, and other fields
• Related Contact’s roles on the deal
• Number of Touchpoints at each stage of the deal - Emails, Chats, Calls, Meetings, Campaigns, Website activity, Documents and Cases.
• Names of documents shared at different stages of the deal

There are many other AI -enabled features that do not require this historical data as explained next.

"79% of opportunity-related data that sales reps gather are never updated in the CRM system." - ESNA

5. What if we do not have data of past deals?

AI based predictions and recommendations get better with more data. But, if you are starting with no data at all, then some of the AI capabilities might not be available to you on Day 1. Typically, at least 200 won/lost deals are required to have sufficient level of confidence.

Following AI-driven features do not need data of historical deals:

Best time to contact

This is derived based on contact’s engagement history (when did contact open email, send email, view document, etc.)

Deal and Contact Sentiment

Sentiment is derived from messages sent by the contact via Email content, Call transcript, & Chat messages.

Deal Conversation Alerts

Sales Managers can be alerted when specific phrases are mentioned by prospect or client.

Conversations Analysis

Specific phrases and keywords can be configured as positive or negative or neutral. Administrators can also put them in different categories (ex: pricing, value proposition, features) and then see a summary of the conversations from the Deal.

Email reply assistant

Sales reps can be provided template and document recommendations when replying to an email based on the information requested by the prospect.

Sales Coaching

Calls and Emails recordings can be analyzed to see how salespersons are communicating. Voice analysis can measure talking speed, mood, Talk2Listen ratio, and advanced insights such as whether a salesperson handled questions well, asked questions, and paused to allow customers to continue speaking.

"30% of all B2B companies will employ AI to augment at least one of their primary sales processes." - Gartner

6. Why do some AI-guided sales initiatives fail?

When embarking on the journey to AI-lighten your team, you should keep in mind that some past attempts by other teams at bringing in AI have failed.

Understanding the reasons for the failure will help you avoid the pitfalls.

So, why do some AI based sales initiatives fail?

• Data Paucity and Poor Quality of data
• Wrong training models employed by the AI engine
• Scattered communications on current deals
• Lack of trust and Poor adoption of AI based features by Salespersons
• Lack of Persistence

"92% of all customer interactions happen on phone" - Salesforce

Data Paucity and Poor Quality of data

To deliver predictions and recommendations, AI engines use data from your past deals. This includes deal related data such as calls, emails, tasks completed, documents sent, contact fields (Ex: Deal Contact Role), org fields (ex: Org industry, location).

With more data, the predictions and recommendations, such as best time to contact, deal score, deal recommendations, get better.

Tip:

If you already have past deal data, then you can expect good results very early in the AI adoption.

Vtiger can also import your past data from other CRMs if your organization is looking to use Vtiger Calculus AI.

Some features of AI, primarily those enabled by Natural Language Processing (NLP), such as Call Analysis, Email analysis, Sentiment score, Conversation Signals, do not need historical data. So, making these as part of your initial goals will help your Salespersons from Day 1.

Note: Vtiger can import your past data from other CRMs.

"Lost productivity and poorly managed leads cost companies at least $1 trillion each year" - CMO Council

Fixed Training Models

Different Machine Learning (ML) models can be applied on historical data to make predictions. The choice of the model can have a big influence on results. ML models that apply self -learning methods deliver more accurate results.

Tip:

Be aware that the training models might need adjustments. Vtiger team is ready to work with you to monitor results and make the necessary adjustments. Vtiger Calculus AI allows administrators to customize the models, and provides easy to configure controls to get the best results for your organization.

Calculus AI also offers salespersons to correct the sentiment with 1 click if they find that a certain text in an email or call recording transcript was wrongly interpreted by the system .

"40% Is the amount of time sales reps spend looking for someone to call" - Inside Sales

Scattered communications and hidden touchpoints

If the calls, emails, chats, WhatsApp conversations, are not logged in the system, then AI engine will offer poor predictions & recommendations, based on partial data. Some touchpoints (ex: engagement on the Quote or ROI document your salesperson sent) might be off the radar but are equally critical to track engagement and arrive at the right prediction.

Tip:

Find a tool that has the plugins and integrations that automatically bring the calls, chats, WhatsApp conversations and emails into the CRM without any effort by Salespersons.

Vtiger CRM Mobile app and the Web client allows salespersons to make calls, and have WhatsApp conversations from within the app. Vtiger also integrates with Zoom Meet & Google Meet. (Microsoft Teams integration is coming in Q1 2021). Vtiger also has add-ons for Gmail and Office365.

Vtiger CRM has built -in document tracking. So, when you send an email from Vtiger with a Quote or any other document, the CRM will not only alert you when they view it, but also use the data to update predictions and recommendations.

"50% Research shows 35 -50% of the sales goes to the vendor that responds first" - Inside Sales

Poor adoption

Inaccurate predictions or recommendations can quickly dampen enthusiasm and reduce adoption. So, it is important to set the right expectations at the beginning and do a feature wise roll out of AI. Since some of the AI based features might need data, it is better to roll those out in the 2nd or 3rd month.

Tip:

Use tools that require minimal change in habits. Set a 2 or 3 phased schedule to roll-out AI. Phase 1 can be features that do not need historical data (noted above).

"85% Or prospects and customers are dissatisfied with their on-phone experience" - Salesforce

Lack of Persistence

As with any new initiative, there will be obstacles. Especially initiatives that require some change in habits, even if minor. You should expect to see challenges when rolling out AI-guided selling.

They might be in the form of inaccurate predictions due to wrong models or insufficient data, or poor adoption due to lack of training.

Knowing that these are par for the course, and moving forward with the corrective actions will lead your Sales team to a successful outcome.

Tip:

Your front-line managers are critical to successfully roll-out AI. They should be part of the planning and monitoring the adoption of the AI feature roll-out.

Share feedback with your Vtiger CRM coaches to get guidance.

"42.5% Of sales reps take 10 months or more to become productive" - Accenture

"Triple-digit growth is expected in areas such as predictive intelligence (118%) and lead-to-cash process automation (115%) in the next three years." - Salesforce

7. What should you look for when selecting a tool for AI-guided selling?

You might not have all the data you need when you begin implementation of AI -guided selling in your organization. So, when looking for the right tool, you should seek tools that offer these capabilities.

• Configurability
• Day 1 features

Configurability

While many tools promise AI capabilities out of the box with little configuration, they either need thousands of records, or deliver inaccurate predictions. With manual intervention, the accuracy of predictions can be improved quickly even with less data. For example, you might have high conversion rates in specific geographic regions. This can be detected only if the region is used in the training model. Some tools allow fine tuning of the fields in the training model.

"2x Chances that high performing companies have sales automation compared to low performing ones" - Velocify

Tools that offer customization deliver better results and increase adoption of AI in your organization.

What you might customize:

• Training models
• Journey Templates (Playbooks)
• NLP Model
• Fields for Fit Score
• Conversation signals - Positive and Negative phrases specific to your business
• Competitor names

Day 1 features

Even if you do not have data of past deals, while waiting for the data to build up, there are ways AI can help salespeople and sales managers on Day 1.

"33% Is the actual time a sales rep spends actively selling" - CSO Insights

Look for tools that can offer these features which don’t need data of historical deals:

• Contact sentiment and Deal sentiment
• Best time to contact
• Action-Items from a call recording
• Coaching dashboards

"High-performing sales teams are 2.8x more likely to be outstanding or very good at predictive intelligence." - Salesforce

8. The path to AI-accelerated success

AI can transform the way your team sells. The actions they take on deals, timing of the follow ups, quality of conversations, documents that are shared will all improve when you empower them with the AI enabled recommendations.

Here are a few simple steps you should take to ensure a successful outcome when you begin rolling out AI in your sales team.

Step 1 - Start with simple objectives

What do you wish to achieve by using AI - Is it to improve lead conversion? To improve the deal win rate? To close deals faster? Or is it to coach sales reps to do better discovery or closing?

Keep in mind that unless the sales reps benefit from it, adoption will be low for any new sales technology. So, one of the primary objectives should be to empower sales reps.

AI can help sales reps in many facets, from prioritizing leads, to identifying actions to be performed on a deal, or alerting them of idle deals, or suggesting content to include when replying to an email, and more.

"High-performing sales teams are 10.5x more likely than underperformers to experience a major positive impact on forecast accuracy when using intelligent capabilities." - Salesforce

Step 2 - Communicate to all the stakeholders

AI guided selling might be received with skepticism by your sales team. We noticed it in our own sales team as well.

Sales persons have 2 primary concerns.

• Accuracy of predictions: AI gets better with more data. So, in early phases there are bound to be instances where the predictions and recommendations are not accurate. Set expectations accordingly.
• Fear that AI might limit their role: On the contrary, AI enhances their success by being a smart assistant for them.

Inform your sales team on what you are looking to achieve by implementing AI guided selling, and how it will help them and the organization.

"46% of companies say that marketing and sales is the area where they are most investing in AI adoption systems." - Salesforce

Step 3 - Find a tool that lets you start with the data you have

AI driven applications require data to train the prediction models. It is not just names & deal size of won and lost deals , but also touch points on the deals, and other attributes of the deal and the related org, and contacts. Usually, this data might not be available if you are starting out afresh.

Tools like Vtiger Calculus can help you get started with limited data, and get you going on Day 1 with features that do not require historical data. For example, your team can benefit from features such as Best time to Contact, Call analysis, and Coaching dashboard, Coaching scorecards 139% out of the box.

"139% is the projected rate of increase in adoption of AI by sales teams in the next three years." - Salesforce

Step 4 - Gather feedback via weekly reviews for the first 90 days.

Gather feedback from the Salespersons and Sales leaders on which features are and are not helping them meet their goals. . AI engines use different algorithmic models on the data to make predictions. These models might need some tweaking to ensure that the predictions and recommendations are accurate. The weekly review will help catch deviations as early as possible and improve accuracy of predictions.

9. Summary

In a recent report titled ‘AI Guided Selling’ Gartner found that COVID-19 epidemic has caught many Sales teams off guard. Revised forecasts are 50% off from previous baseline forecasts. Many Sales leaders are now looking to modernize internal and external processes to improve conversions and accuracy in forecasts.

By utilizing AI powered tools, sales leaders can provide an able assistant that will guide their salespersons at all steps of their journey with each customer. They also can transform to become mentors.

"81% of Fortune 500 CEOs consider AI a crucial area to invest." - Forbes

"83% of the most aggressive adopters of AI and cognitive technologies said their companies have already achieved either moderate (53%) or substantial (30%) benefits." - Deloitte