The short answer: Courted's Potential to Move score estimates how likely a real estate agent is to change brokerages within the next 12 months. Potential to Grow estimates how that agent's production is likely to change over the next 12 months. One is primarily a timing signal. The other is a trajectory signal.
That distinction matters.
An agent can be likely to move without being likely to grow. Another agent can be on the verge of a breakout year but have little reason to leave their current brokerage. The most compelling recruiting targets are often agents who score highly on both dimensions, especially when there is also a warm connection between the agent and your organization.
The scores are not crystal balls, and they are not intended to replace a brokerage leader's judgment. They are prioritization tools. Their value comes from helping a recruiting team decide where to focus its next call, message, meeting, or relationship-building effort.
What is Courted's Potential to Move score?
Courted's Potential to Move score forecasts an agent's likelihood of changing brokerages over the next 12 months.
The model was developed by examining approximately 15 years of historical agent data. Courted identified agents who changed brokerages, compared them with agents who stayed, and analyzed more than 1,000 features associated with those outcomes. Those patterns are then used to estimate the relative move potential of agents today.
In practical terms, the score asks:
Based on the behavior of similar agents in similar circumstances, how unusual would it be if this agent changed brokerages during the next year?
Agents categorized as having high Potential to Move are about three times more likely to change brokerages than the average agent. Courted has also cited model results in which roughly four out of five agents who moved had been identified by the model.
Those statistics describe two different things. The first measures how concentrated movers are in the high-score population relative to the market baseline. The second measures how many eventual movers the model surfaced. Neither means that every high-score agent has an 80% individual probability of moving.
That is an important piece of probabilistic hygiene. A useful forecast does not need to be certain. It needs to rank opportunities better than the alternative.
What factors can indicate that an agent may move?
Courted evaluates a broad set of historical, production, activity, career, and market characteristics. Signals associated with agent mobility can include:
- A prolonged period without an active listing, pending deal, or closing
- A sustained decline in transaction count or sales volume
- A sharp drop from the agent's previous production range
- Short tenure at prior brokerages or a history of frequent moves
- A mismatch between strong historical production and a recent slowdown
- Changes in office leadership or managing broker structure
- Elevated agent departures or instability within the current office
- Recent co-brokered transactions or other relationships that create a warmer path to outreach
No single factor determines the score. An agent with no recent closings is not automatically preparing to leave. The model's purpose is to evaluate the combined pattern, not to turn one data point into a verdict.
This is also what the score does not use: it does not monitor an agent's browsing history, social media viewing behavior, private communications, or Courted customer usage data. It is based on observable business characteristics and historical patterns.
How should a brokerage use Potential to Move?
The simplest use case is recruiting prioritization.
If a brokerage has 2,000 possible recruits in its market and time to contact only 20 today, a ranked list is more useful than an unfiltered database. The best near-term candidates often combine:
- High Potential to Move
- No active or pending business
- A recent decline from a stronger production history
- A relevant production range for the brokerage
- A mutual relationship, co-broke, or other credible reason to connect
The score should guide the order of operations, not dictate the message. A high score does not justify sending a generic "ready to leave?" email. It creates a reason to research the agent and begin a relevant conversation.
For example, a brokerage might recognize a recent listing, congratulate the agent on a strong year, reference a mutual connection, invite the agent to an event, or offer insight on an area of the agent's business. The prediction helps select the audience. The recruiter still has to earn the response.
What is Courted's Potential to Grow score?
Courted's Potential to Grow score estimates how an agent's production is likely to change over the next 12 months.
The underlying question is different from mobility:
Is this agent's business likely to expand, remain stable, or contract relative to comparable agents?
Courted analyzes historical agent businesses over a roughly 15-year period and identifies what future high performers tended to look like earlier in their development. Inputs can include average sales price, transaction activity, days on market, listing and pending trends, co-brokerage patterns, team relationships, and multi-year production trajectory.
The model produces a predicted sales-volume range for the next 12 months, including a 90% confidence interval. Courted then evaluates the projected change relative to the agent's last 12 months and benchmarks the result against similar agents. Agents are categorized as:
- Limited Growth
- Stable
- Moderate Growth
- High Growth
The confidence interval matters because a forecast should express uncertainty rather than hide it. The point is not to claim that an agent will produce exactly $8.4 million next year. The point is to estimate a realistic range and determine whether the available evidence points upward, sideways, or downward.
What factors can indicate that an agent may grow?
Signals associated with future production growth can include:
- Increasing transaction count or closed sales volume
- More active listings and pending transactions
- Gains in local market share
- Improving sales velocity or listing performance
- A growing network of co-brokers and collaborators
- Repeated work with higher-producing agents
- Emerging team-leadership or co-listing patterns
- A steady multi-year upward trajectory rather than a one-year spike
Again, none of these is determinative. A large transaction can temporarily distort an agent's volume. A changing market can raise or lower nearly everyone at once. Benchmarking an agent against similar professionals helps distinguish a real trajectory from background noise.
Potential to Move vs. Potential to Grow
The easiest way to understand the difference is to ask what decision each score improves.
ScorePrimary questionBest useTypical time horizonPotential to MoveWho may be approaching a brokerage change?Prioritizing near-term recruiting and retention outreachNext 12 monthsPotential to GrowWhose production is positioned to increase?Finding rising stars, evaluating recruiting upside, and prioritizing coachingNext 12 months
Potential to Move is mainly about timing. Potential to Grow is mainly about trajectory.
The inputs can overlap because an agent's business circumstances can affect both outcomes. A production decline may coincide with a greater willingness to consider a change. Rising listings and pendings may indicate future growth. But the same signal can mean different things in different contexts, which is why the two predictions should not be collapsed into one score.
How should teams interpret the two scores together?
Combining the scores creates four distinct recruiting and retention profiles.
Potential to MovePotential to GrowRecommended interpretationHighHighPriority recruit. The agent may be open to change and could bring meaningful future upside.HighLowNear-term recruiting opportunity, but evaluate strategic fit and expected production carefully.LowHighStrong retention or development priority if the agent is already at your brokerage. If external, nurture the relationship over time.LowLowLower immediate priority unless there is a strong relationship or another strategic reason to engage.
The most attractive recruiting profile is usually High Move + High Growth + Warm Connection.
But even this combination is not a command to recruit. Compensation preferences, culture, geography, team structure, personal circumstances, and the quality of your value proposition still matter. A model ranks the probability of an outcome under uncertainty. It does not conduct the recruiting conversation for you.
How accurate are real estate recruiting predictions?
Accuracy depends on what is being measured.
Courted has reported that approximately four out of five agents who changed brokerages were predicted by its model, and that agents with high Potential to Move are about three times more likely to move than the average agent.
Brokerages should not interpret those statements as a guarantee that four out of every five high-score agents will move. The base rate matters. If brokerage changes are relatively uncommon, even a group that is three times more likely to move will still contain many agents who stay.
That does not make the prediction weak. In recruiting, the relevant comparison is usually not certainty. It is the performance of a prioritized list compared with random prospecting, production-only lists, or recruiter intuition alone.
Field examples have shown recruiting conversion rates ranging from roughly 2.5% to more than 6%, depending on market, targeting, cadence, coaching, and outreach quality. In some cases, a single successful recruit can justify the annual cost of the platform. Those outcomes are examples, not guaranteed results.
What is the best workflow for using predictive analytics in agent recruiting?
A practical workflow looks like this:
- Define the recruit profile. Select the production range, geography, specialty, and experience level that fit your brokerage.
- Prioritize by Potential to Move. Start with agents most likely to be approaching a decision window.
- Layer in Potential to Grow. Favor agents whose future trajectory aligns with your recruiting goals.
- Add current business context. Review active listings, pendings, recent closings, production change, office movement, and tenure.
- Find the warmest path. Look for mutual relationships, recent co-brokes, shared markets, or relevant accomplishments.
- Take consistent, personal action. A focused routine of 10 to 20 targeted actions per day, including three to five personal calls or messages, can create more reliable pipeline movement than occasional mass outreach.
- Track results and recalibrate. Measure response, meeting, opportunity, and recruited-agent conversion rates by score, segment, message, and recruiter.
This last step is easy to overlook. A prediction can tell you who deserves attention. Your own conversion data tells you how effectively your organization turns attention into appointments and appointments into recruits.
When should you override the score?
Whenever you have better direct information.
Suppose you met an agent for coffee yesterday and they gave you a credible, specific reason they will not move. The model may still classify them as high Potential to Move because their business profile resembles agents who have moved historically. Your firsthand knowledge should take precedence.
The reverse is also true. A low-score agent may privately tell you that they are unhappy with their brokerage. That information is more current and specific than a statistical forecast.
Good predictive systems combine three layers:
- The model's broad pattern recognition
- The recruiter's local knowledge and relationships
- The agent's current, expressed needs
The model is strongest at narrowing the field. Humans are strongest at understanding the exception.
Limitations to keep in mind
Predictive analytics works best when the underlying data is complete and the target behavior resembles the historical behavior used to train the model.
Results can be affected by:
- Limited or incomplete MLS coverage
- Unusual local-market dynamics
- Career breaks, relocations, or personal decisions not visible in production data
- Sudden brokerage changes that have little historical precedent
- Individual circumstances that no market-level dataset can observe
This is why Courted presents forecasts as probabilities and ranges, not certainties. The model is designed to improve prioritization, not eliminate uncertainty.
Frequently asked questions
What is a likelihood-to-move score for a real estate agent?
A likelihood-to-move score estimates the probability that an agent will change brokerages within a defined period. Courted's Potential to Move score forecasts brokerage mobility over the next 12 months using historical business, production, activity, career, and market patterns.
What is a likelihood-to-grow score for a real estate agent?
A likelihood-to-grow score estimates whether an agent's production is likely to increase. Courted's Potential to Grow model predicts a sales-volume range for the next 12 months, benchmarks the projected change against similar agents, and classifies growth potential as Limited, Stable, Moderate, or High.
What is the difference between Potential to Move and Potential to Grow?
Potential to Move estimates the timing of a possible brokerage change. Potential to Grow estimates the direction and magnitude of an agent's future production. The first helps teams decide who to contact now. The second helps them decide who may create the most future value.
What is the best predictive profile for an agent recruit?
The strongest general profile is an agent with high Potential to Move, high Potential to Grow, and a warm connection to your brokerage. Teams should also consider production fit, current activity, office context, geography, and the agent's individual goals.
Does a high Potential to Move score mean an agent will definitely switch brokerages?
No. It means the agent has characteristics historically associated with a higher rate of brokerage movement. The score should determine outreach priority, not be treated as proof of intent.
Can brokerages use Potential to Grow for retention?
Yes. A current agent with high growth potential may deserve proactive coaching, recognition, lead support, or a career-development conversation. Predictive analytics can help brokerages protect rising talent as well as recruit it.
How much daily activity is needed to turn predictions into recruiting results?
The right volume varies by brokerage, but teams often build consistency with 10 to 20 focused recruiting actions per day, including three to five highly personalized calls or messages. Targeting improves the list, but relevance, cadence, and follow-through determine conversion.
The bottom line
Most brokerage recruiting problems are not caused by a complete lack of possible candidates. They are caused by an excess of candidates and a shortage of time.
Courted's Potential to Move score helps answer, "Who may be entering a decision window?" Potential to Grow helps answer, "Who is positioned to become more valuable over time?" Used together, they help brokerage leaders spend more of their recruiting effort on agents with the right timing, trajectory, and fit.
That is the practical promise of predictive analytics. It does not make recruiting certain. It makes recruiting less random.
