How can AI transform mid-funnel sales activities compared to traditional approaches?

AI shifts mid-funnel sales from reactive and manual to proactive and data-driven – it provides real-time deal health and risk detection, recommends precise next actions, and predicts outcomes much more accurately than traditional, manual processes.

It can help teams improve focus, consistency of pipeline, and forecast reliability by:

Automating administrative tasks (like data entry and inspection)
Surfacing insights within existing workflows
Keeping managers and reps aligned on priorities

What steps should organizations follow to successfully implement AI-driven mid-funnel solutions?

Step 1: Integrate all relevant data sources – including your CRM, email, meetings, etc. Start with the basics to establish a solid foundation.

Step 2: Clean and standardize data – clean your accounts, contacts, opportunity fields, stages, activity records, etc.

Step 3: Embed AI insights (deal scores, suggested actions, forecasts) directly within the workflows and tools reps and managers already use – to minimise disruptions and change management, while ensuring quick wins.

Step 4: Educate and enable teams for adoption – explain how AI insights work and ensure there’s a feedback loop (ideally through AI coaching).

Step 5: Assign clear ownership – across senior leadership (typically champions), Revenue Ops (process and tech), enablement (sales capability), and continuously monitor results and iterate. AI is a living and breathing system.

What are the different uses for AI in the middle of the funnel?

Deal Intelligence & Analytics:

Deal scoring – AI flags a deal’s likelihood to close, identifies risks, and scores opportunity health.
Pipeline & AI Deal Inspection – AI scans all deals for risk factors, gaps, missing steps, or exceptions to your defined process.
Predictive forecasting – more accurate, dynamic pipeline predictions and scenario modeling.
Conversation and sentiment analysis – extracting buyer intent, risk, and opportunity signals. Powered by tools like Clari Copilot, Gong, Chorus (ZoomInfo), Avoma, and Outreach.
Multi-threaded stakeholder mapping – showing engagement gaps and prompting outreach to the right stakeholders. Powered by tools like Clari Relationship Intelligence, Mutual Action Plans, and Reveal AI.

Process Automation & Compliance:

Pipeline/deal inspection automation – flagging gaps, missing steps, and noncompliance. Powered by tools like Clari AI Agents, Gong Forecast, Salesloft Rhythm, and Salesforce Einstein.
Automated data entry and CRM hygiene – reducing manual admin effort, all activities can now be logged automatically. Powered by tools like People.ai, Clari, Outreach, Salesforce AI, and Einstein Activity Capture.
Sales process compliance tracking – ensuring methodology adherence and nudging, e.g., MEDDIC. Powered by tools like Clari, InsightSquared, Salesforce, and Mediafly.

Actionable Guidance & Workflow Optimization

Next-best-action recommendations – e.g., who to contact, follow-up timing, and coaching.
Emerging agentic workflows – autonomous AI agents that orchestrate pipeline actions and updates.
Rep and Manager Coaching Insights – analyzes sales calls, emails, and meeting transcripts for behaviours correlated with wins/losses (talk-to-listen ratios, objection handling, competitor mentions). Coach based on person-level data for higher impact. Powered by tools like Gong Coach, Chorus Coaching, and Clari Manager.
Content & Collateral Optimization – tracks which playbooks, decks, or emails are most effective at different sales stages or for specific personas for effective pipeline generation. Powered by tools like Highspot, Seismic, and Showpad.

Deal Scoring

What is AI-enabled deal scoring? How do you set it up?

Deal Scoring: What it is

Deal Health & Risk Identification – AI analyzes real-time and historical engagement signals (e.g., emails, calls, meeting participation, CRM data, product usage) to score deals on likelihood to close. It flags stalled, at-risk, or high-potential deals based on actual behavior – not just static CRM fields or gutfeel notes from reps.

Automated Insights for Managers – managers get real-time visibility into pipeline health. AI surfaces coaching opportunities, performance gaps, and process adherence issues. For example, “this deal is in stage 4 and commit, but no activities in the last 14 days, and no Economic Buyer has been identified.”

Productivity Improvements – by eliminating manual data entry and automating deal inspection, AI frees up significant time for both reps and managers. Teams spend less time getting the context and more time executing and selling.

How to set up AI Deal Scoring

Step 1: Data integration: Connect CRM (e.g., Salesforce), email, meeting, and other engagement systems, such as data providers.

Auto-capture activities across all relevant channels: Emails, meetings, calls, and more.

Step 2: Ensure data structure & quality:

Ensure your core revenue data is structured, e.g., accounts, contacts, opportunities, etc.
Standardize contact journey, accounts, deal stages, opportunity fields, and engagement signals for model training.

Step 3: Prompt engineering/configuration: There are two types of prompts

Automated prompts: This gives your standard information automatically without the need for prompting because you’ll likely need it, e.g,. account, opportunity, and call overviews
Manual prompts: If you’re interested in specific areas of the deals, you can either a) prompt via a dropdown menu (e.g., “competitors”, follow-up email”, etc.) or insert your own prompt (e.,g. “what are the project stages?”

Step 4: Establish Workflow/Revenue Cadence:

Embed deal scoring and AI insights directly into established revenue cadences – pipeline reviews, major deal reviews, rep 1:1s – from the seller to the CRO
Automate risk flagging and recommendations to trigger follow-ups, manager interventions, or workflow actions in real time.
Continuous learning from your historical data and deal outcomes allows for contextual scoring and recommendations to improve.

Note: The process depends on your tech stack – the above process details the main systems involved in a typical tech stack.

What needs to be established across your Sales process and tooling in order to build AI Deal Scoring?

Deal Scoring: Process & Tooling
Sales Process Considerations	Maintain high-quality, up-to-date opportunity data – “Bad data = bad outcomes” – AI requires trusted input to deliver credible deal scores (see above). You need at least 4 core sales processes: 1. Sales methodology: For example, Force’s Command of the Message Methodology to align with the buyer’s need. 2. Qualification Methodology: For example, the MEDDICC framework for qualifying deals 3. Sales Stages: Map out your Sales methodology and Qualification Methodology to 4-5 critical sales stages and required actions so AI can identify compliance gaps and recommend corrective actions. 4. Operational Cadence: This answers the questions “who does what, when, why, how, and what sales impact is that driving?” AI is most effective when aligned with structured, repeatable Operational Cadence (e.g., weekly pipeline reviews and forecast calls, monthly business reviews). This ensures that insights are actionable and tied to business outcomes. Ensure team training and adoption – especially for interpreting AI scores and acting on prescribed next steps. Managers and sellers should understand how AI scores are generated, what drives risk, and how to leverage recommendations in their daily workflows.
Available tools	Most important is that you identify your focus areas, then align your workflows with the tech. – there are many tools available, and they depend on the growth driver you’re focused on. Revenue tech can be split into 3 areas: • Pipeline generation: Outreach, Salesloft, Clari, Apollo, etc. • Deal Management: Clari, Gong, Outreach, Pipedrive, etc. • Pipeline and Forecasting: Clari, Gong, Outreach to some extent *Besides these, there are supporting tech,* for example, Data Providers such as Clay, LinkedIn, ZoomInfo, Apollo, etc.

Suggested Actions

What are AI-enabled suggested actions? How do you set up a process for AI-suggested actions?

Suggested Actions: What it is

AI continually analyzes deal activity, engagement signals, and historical outcomes to recommend what action a seller or manager should take next – for example, it could tell a seller to schedule a follow-up, engage new stakeholders, or address an objection in a specific way. AI highlights which actions will have the most impact, helping reps focus on value-adding activities and avoid “random acts of selling.”

It helps to standardize best practices – like when and who to loop in, and solutioning recommendations. AI helps ensure no key steps are missed, reducing the number of slipped or stalled deals.

AI surfaces coaching opportunities for managers – For example, it might suggest a manager join an at-risk account call or prompt a forecast adjustment based on recent signals – ensuring managers are not flying blind but can make changes before it’s too late.

It automates action nudges – action recommendations can appear as in-app prompts, email notifications, workflow tasks, or even automated calendar invites – making next steps clear to act on.

How to set up AI Suggested Actions

Step one: Build data & integration foundations – integrate your CRM (e.g., Salesforce), Revenue Platforms, and communication channels (email, calendar, calls, video meetings). Ensure timely, bi-directional sync so AI models receive up-to-date signals and can push recommendations into the user’s natural workflow.

Step two: Structure data fields and prepare recommendation logic – standardize opportunity and activity fields (e.g., deal stages, last engagement date, roles of buying team members). Configure your Revenue Platform key actions and AI recommendations (pipe gen, deal management, forecasting).

Step three: Implement recommendations across the CRM or Revenue Platform – deliver AI-driven recommendations through the tools reps and managers use most, mainly the CRM or Revenue Platform. Integrate Suggested Actions with pipeline and forecast reviews, ensuring both reps and managers use these signals to guide deal strategy and meeting agendas. Enable “closed loop” feedback collection (e.g., reps mark an action as completed or not relevant), which continually tunes the AI.

What do you need to establish across your Sales process and tooling in order to incorporate AI-suggested actions?

Suggested Actions: Process & Tooling
Sales Process Considerations	Clearly define what “good” looks like at each funnel stage – refer to the advice on this in sales process considerations for deal scoring. This lets AI align recommendations to real business objectives and sales growth. Drive rep and manager buy-in through education and visibility into how AI derives its recommendations – transparency is key; users need to understand the “why” behind the transformation. Integrate new behaviors gradually, celebrate early successes, and use champions to showcase impact and reinforce adoption across the sales team. Embed recommendations into existing sales processes, including: • Pipeline reviews • Deal inspection meetings • Rep 1:1s Encourage users to provide feedback on the quality and usefulness of suggestions – this helps calibrate the AI to your evolving sales process and language. Data Governance – maintain high-quality, up-to-date activity and opportunity data. Incomplete or inaccurate data undermines AI effectiveness and user trust.
Available tools	*Because “Suggested Actions” are tied to “Deal Management,” the same tools are applied as they have the data to provide Next Best Actions.*

Predictive Forecasting

What is AI Predictive Forecasting? How do you set up Predictive Forecasting?

Predictive Forecasting: What it is

It provides current quarter, accurate pipeline predictions- AI algorithms uncover patterns and signals in current opportunities – engagement, deal history, activity trends, and external factors – to increase forecast accuracy beyond what’s possible with spreadsheets or manual rollups. Typically to ~10% of where you’ll land by week 3-5 of the 13-week quarter.

It can extend planning and forecasting into future quarters – the most predictable indicator of whether you’ll miss, hit, or beat your quarter is how much pipeline you started the quarter with. AI allows you to plan ahead so you know how much pipeline, account coverage, and seller activity you need to build the right pipeline coverage for the next quarter. This prevents sales leaders from coming into the quarter already being behind.

Key forecast features include:
• Dynamic risk & upside visibility – AI continuously updates predictions as new data streams in, surfacing likely slippage, identifying deals trending above target, or pinpointing “hidden” risks or upside in the committed forecast.
• Scenario modelling & forecast adjustment – users can run “what-if” analyses, simulate best/worst-case scenarios, or model the impact of major deals slipping or closing early, all powered by AI-driven Path to Plan Methods.
• Forecast granularity & segmentation – AI can forecast by rep, region, product line, segment, etc., detecting micro-trends and variances. This provides actionable, tailored insight for managers.
• Anomaly and outlier detection – the system flags entries that deviate from historical norms or expected behaviours (ex, forecasted deals with no activity in 30+ days), enabling faster corrective action.
• Automated roll-ups & executive summaries – AI produces aligned and transparent top-down, bottom-up, and AI-driven roll-ups, auto-generates summaries, flags variances, and supplies recommended forecast adjustments.
• Uncover hidden revenue – many companies have a very hard time predicting how much revenue will open and close within this quarter. AI helps gain visibility into this revenue stream by evaluating your historicals with deal activities to pinpoint in-quarter revenue.

How to set up AI Predictive Forecasting

Step one: Data integration – connect your CRM systems (e.g., Salesforce), revenue platform, and activity data (emails, meetings, calls).

Step Two: Ensure data quality and preparation:

Ensure historical win/loss and pipeline movement data is accurate, time-stamped, and standardized for AI to ingest and interpret the data accurately.
Define business logic for forecast categories (Commit, Best Case, Pipeline, Omitted, etc.) to align model output to internal language.
Clean up opportunity stage definitions, owner fields, amount fields, and your role/org. Hierarchy, and ensure completeness of required sales data.

Step three: Embed tooling into the daily/weekly sales process:

Serve forecast insights directly inside your revenue platforms, or via tailored forecast workspaces
Set up alerting/notifications for forecast misses, risk detection, and upside signals.
Enable drill-downs from summary to underlying deal-level signals so managers and reps can take immediate action.

What do you need to establish across your Sales process and tooling in order to incorporate AI Predictive Forecasting?

Suggested Actions: Process & Tooling
Sales Process Considerations	Consistent Opportunity Management: • Require reps to maintain up-to-date, accurate opportunity data (stage, close date, amount, probability) as a foundation for AI modeling. Defined Forecast Cadence: • Establish a regular forecast submission and review cadence (weekly, biweekly) across teams, using AI insights as key input along with seller judgment and context Change Management & Training: • Train sales teams and managers on interpreting AI-driven forecasts, variance flags, and using forecast scenario tools. • Create smaller Expert Groups focusing on pipeline, deal management, and forecasting across acquisition, retention, and renewal. Their responsibility is to roll out, train, and validate the impact – should be a cross-functional team of sales, marketing, enablement, and RevOps.
Available tools	Forecasting platforms by tier: • Leader: Clari – deep pipeline analytics, AI-powered forecast calls, scenario modeling, anomaly and risk detection, and advanced integration with CRM/activity signals. • Up and Coming: Gong Forecast – AI-driven forecast submissions, pipeline risk analysis using conversation intelligence, and activity signal incorporation. • Legacy: Salesforce Einstein Forecasting – uses multi-factor historical analysis to deliver AI-driven forecast predictions and recommendations, native to Salesforce. • Second-tier Basic Tools: Aviso, Outreach, Salesloft – core array of tools for AI-based forecast rollups, pipeline inspection, basic analytics, and sales process compliance. • The others: InsightSquared (Mediafly), People.ai, Revenue.io, and others – these platforms increasingly embed forecasting, deal health, and pipeline predictability features using AI models. They’re cheaper solutions and not a robust deal and forecasting platform.

What might future AI developments enable in AI-driven sales?

AI SDRs are already en route – these have been rolled out with the likes of Artisan.co. However, these haven’t been proven and have had some negative impact, e.g., companies burning their accounts because of the low-quality outreach and negative customer experience. Although it’s still early for AI SDRs, they will improve over the next 2-6 quarters to provide satisfactory customer experiences.

Autonomous AI Agents:

Deal Inspection Agents – fully autonomous AI “copilots” that continuously scan opportunities, identify gaps, and take direct follow-up actions (e.g., create tasks for reps, suggest emails, escalate flagged risks)
Meeting Assistants – join calls to take notes, summarize discussions, extract next steps, and update CRM fields automatically, all while the seller is in the call, e.g., with live battlecards
Orchestration Agents – proactively coordinate multi-stakeholder outreach, recommend sequences, and even send nudges or schedule meetings on behalf of users.
Forecast Agents – forecasts on behalf of the rep and managers with explanation and data. Seller and manager can override or accept the forecast. The agent continuously adjusts and improves.

How do you ensure data quality and integration between AI systems and existing sales tools?

Establish a single “source of truth” for data and sales objects – including account, opportunity, people, tasks, fields, etc. Avoid duplicated or siloed data feeds.

Integrate the primary systems – including your CRM (e.g., Salesforce), Revenue Platform, email, calendar, and call/meeting platforms via APIs. Usually, the Revenue Platform has the logic embedded to be able to a) pull in data, structure, interpret, and present it.

What best practices have you seen for integrating AI into the daily workflow of sales teams at the mid-funnel stage? How do you drive adoption and ensure that insights lead to action?

Embed insights where work happens – Surface AI insights and recommendations directly within sales reps’ and managers’ core tools (CRM dashboards, pipeline pages, inbox, meeting summaries). Avoid requiring users to log into standalone AI “dashboards” outside normal workflow; activities and insights should be in the same place and workflow.

Tie AI insights to existing cadence – align deal scoring, suggested actions, forecast signals, etc., with regular pipeline reviews, forecast calls, and 1:1s. Use AI-generated risk summaries to shape deal strategy discussions.

Establish feedback loops – let reps/managers flag AI insights as “useful,” “not relevant,” or “not actionable.” Feed this back into model refinement for a better fit to your business.

Enable champion-driven adoption – enlist early adopters as success champions – have them demo value in team meetings, or share “save stories.” Offer quick-win use cases (e.g., deal risk flag saves a potentially lost deal) to showcase immediate value.

Overall

What are the most important things to get right?

Feed your workflows quality data – without structured data, AI will struggle, e.g., too many account duplicates make it hard for AI to associate tasks and context to an account.

Nail down your sales process – creating a clear sales process that’s defined, communicated, and fits with your sales strategy.

Ensure you have the right talent and culture to embrace with your AI tooling – make AI a high-priority and ensure the right people are aligned (senior leadership, revops, enablement, managers, and reps).

Tech is not a silver bullet but a catalyst for the change you’re driving – ensure the tech you’re implementing also aligns with your strategy, process, and outcomes you’re driving.

Leveraging AI in the Middle of the Funnel