Scoring Mechanics

Copyright DotAlign 2015-2020. US Patent #9,785,781 and patents pending.

Scores with a person

Person-to-person

The person-to-person score represents the relationship strength between a DotAlign data contributor and another person.

Every person found in the mailbox will have a score of at least 1 point. This score will increase as additional activity occurs, and will decrease as the relationship becomes stale, i.e. as time passes without further interactions. The maximum possible score is 100 and reflects an active, current relationship.

To calculate the person-to-person score, the system takes the following inputs:

  • Number of interactions involving the two people, including events and reciprocated emails
  • The date of either (i) the last event, or (ii) the earlier of the last inbound email and last outbound email, whichever is most recent

The score is then calculated as follows:

MAX(1, (MIN(100, InteractionPoints) * RecencyMultiplier)))

Variable

Description

Calculation

InteractionPoints

The raw number of points associated with the relationship. Each relationship gets:

- 1 point per reciprocated email, measured as the lesser of outbound (from contributor) and inbound (from contact) emails

- 3 points per eligible event (e.g. future meetings don't count until they have taken place)

1 + EventPoints + EmailPoints

RecencyMultiplier

The multiplier that decays the relationship score as time goes by without further interactions. The multiplier considers:

- a primary decay, which decays 75% of the relevant points to 0, linearly over 2 years

- a less aggressive secondary decay, which decays 25% of the relevant points to 0, linearly over 6 years

Both the primary and secondary decay factor in how many days have passed since the last interaction.

1 - (PrimaryDecay + SecondaryDecay)

Example
  1. Susan meets Andrew at a conference and sends him an email. Their score is 1. After Susan sends another follow-up email, Andrew replies, increasing the score to 2.
  2. They exchange hundreds of emails, earning a score of 100. After a month of no interactions, their score will be 97. After six months, it will be 79; after one year, 58.
  3. Andrew emails Susan to tell her he is taking a new job, and transitions her to a new account manager. After Susan replies, rekindling the connection, the score is again 100.
  4. Andrew and Susan fall out of touch. After two years, the score is 17. The score continues to decay until it reaches 1, after six years of inactivity between Susan and Andrew.

Group-to-person

The group-to-person score represents the relationship strength between a group of DotAlign data contributors and another person. It is represented as a number between 1 and 100.

To calculate the group-to-person score, the system takes the following inputs:

  • All person-to-person scores between data contributors in the group and the specified person

The score is calculated as follows:

MAX(PersonScores)

Example
  1. Rob and Brian have a relationship score of 43. Jennifer knows Brian to the tune of a 71, and Elijah and Brian know each other at a score of 16.
  2. As a group, Rob, Jennifer, and Elijah have a score with Brian of 71.

Scores with a company

A score with a company is a measure of how much influence a relationship might have with a company, as based on the influence via each contact associated with the company.

Influence score (component)

The influence score calculates the impact of a data contributor's relationship with a contact at a company on the company-level scores, and is a key building block of those top-level company scores.

This score scales down the impact that any person-to-person relationship should have on a company, based on seniority. As time passes without further evidence that the contact is still employed at the company (or once the contact is believed to have left, of course), the influence score will decrease. Even after many years, the influence score will still be at least 1.

To calculate the influence score, the system takes the following inputs:

  • The person-to-person score the data contributor has with the contact at the firm
  • The seniority level of the contact at that firm
  • Whether the contact is believed to have left the firm or not
  • How long it has been since anyone saw evidence of the contact being at the firm

The score is then calculated as follows:

MAX(1, (PersonalScore * SeniorityMultiplier * RecencyMultiplier))

Variable

Description

Calculation

PersonalScore

The person-to-person score between the data contributor and the contact.

See person-to-person section

SeniorityMultiplier

The scaling factor that determines how much influence a given relationship has with the company. Senior relationships have more influence, all else equal.

This factor effectively caps influence at 70.

0.7 (contacts who are SVP-level+)

OR

0.5 (others)

RecencyMultiplier

The multiplier that decays the influence score as time goes by since the contact was last associated with the company. The multiplier considers:

- whether the contact has left the firm or not

- a primary decay, which decays 75% of the relevant points to 0, linearly over 2 years

- a less aggressive secondary decay, which decays 25% of the relevant points to 0, linearly over 6 years

For former employees, decay begins on the day following the last evidence date. Current employees get a grace period of 1.25 years before decay begins.

1 - (PrimaryDecay + SecondaryDecay)

Example
  1. Diana knows Charles, a freelance journalist, very well; they have a score of 100. As Charles is a freelancer and is not associated with any company, no influence score will be computed.
  2. Charles receives an offer to join a publisher as a Vice President and emails Diana from his new job. Diana's relationship with Charles has an influence score of 50, within the context of the publisher.
  3. Several months pass and Diana emails Charles to see how his new job is going. Charles replies, telling Diana he has subsequently been promoted to SVP and now runs his division. The influence score increases to 70.
  4. A year and a half goes by without any further interaction. Their personal score has decreased to 38. This, combined with the elapsed time exceeding the influence decay grace period, reduces the influence score to 24.
  5. Diana finds out that Charles has started his own firm, and updates her contact card for him. Diana's relationship with Charles results in an influence score of 27 in the context of his current firm. The influence score in the context of his former employer, the publisher, drops to 10.
  6. Diana sends Charles a congratulatory email. After he responds warmly and suggests they catch up soon, their personal score is again 100. The influence score related to the new firm increases to 70, and the score related to the publisher increases to 26.

Person-to-company

The person-to-company score represents the relationship strength between a DotAlign data contributor and a company, via all the contacts the contributor knows at that company.

This score will increase as the data contributor nurtures relationships with more people associated with the company, especially if those contacts are senior. The score will decrease as contacts leave the firm or as it becomes less likely that they are still employed at the firm. The score is between 1 and 100.

To calculate the person-to-company score, the system takes the following inputs:

  • The influence score the data contributor has via each contact at the firm

The score is calculated as follows:

MAX(1, ((MainContactInfluence + OtherContactsInfluence) * InfluenceMultiplier)

Variable

Description

Calculation

MainContactInfluence

The contributor's main relationship at the firm counts for 80% of their overall score with the company, and factors in:

- The highest influence score the contributor has with a contact at the firm

0.8 * MAX(InfluenceScores)

OtherContactsInfluence

The contributor's other relationships with the firm count for 20% of their overall score with the company, and factors in:

- The sum of all influence scores the contributor has with the company, excluding the highest influence score

0.2 * (SUM(InfluenceScores) - MAX(influenceScores))

InfluenceMultiplier

The multiplier exists to make sure there is a limit on the effect of having only many weak relationships, and will be either:

- 1, if the contributor has at least one strong, active relationship with a senior contact at the firm (i.e. if the highest influence score is above 50)

- 0.85, if the contributor has no such influential relationships

1 (if 1+ influential relationships)

OR

0.85 (no influential relationships)

Example
  1. Francesca and Harold, a bank Vice President, know each other very well and have a score of 100. The influence score for their relationship in the context of that bank is 50. Harold is the only person Francesca knows at the bank, giving Francesca a score of 34 with the bank overall.
  2. Francesca maintains her relationship with Harold and also earns a score of 100 with Ian and Kevin, two senior associates on Harold's team. Those relationship also yield an influence score of 50. Francesca now has a score of 51 with the bank.
  3. Harold makes Managing Director, increasing the associated influence score to 70. That, combined with the fact that Francesca now has a strong senior-level relationship at the bank, increases Francesca's score with the bank to 76.
  4. Harold introduces Francesca to his colleague, Georgina, who is also an MD. They form a similarly strong relationship, boosting Francesca's score with the bank to 90.
  5. Harold leaves the bank. Francesca maintains her relationship with him as well as with Georgina, Ian, and Kevin. After one year, her relationship with the bank is 84. After two years, it drops further to 78 - a little bit higher than if she knew only Georgina, Ian, and Kevin. After a total of six years after his departure, his relevant influence will have decayed fully to a 1, and her score with the bank would be 76.

Group-to-company

The group-to-company score represents the relationship strength between a group of DotAlign data contributors and a company, via all the contacts the group of contributors know at that company.

Like the person-to-company score, this score will increase as the data contributors nurture relationships with more people associated with the company, especially if those contacts are senior. The score will decrease as contacts leave the firm or as it becomes less likely that they are still employed at the firm. The score is between 1 and 100.

To calculate the group-to-company score, the system takes the following inputs:

  • For each contact at the firm, the maximum influence score any contributor has with them

The score is calculated as follows:

MAX(1, ((MainContactInfluence + OtherContactsInfluence) * InfluenceMultiplier)

Variable

Description

Calculation

MainContactInfluence

The group's main relationship at the firm counts for 80% of their overall score with the company, and factors in:

- The highest influence score any contributor has with any contact at the firm

0.8 * Max(MaxInfluenceScores)

OtherContactsInfluence

The groups's other relationships with the firm count for 20% of their overall score with the company, and factors in:

- For each contact at the firm (excluding the main contact), the maximum influence score any contributor has with them

0.2 * (SUM(MaxInfluenceScores) - MAX(MaxInfluenceScores))

InfluenceMultiplier

The multiplier exists to make sure there is a limit on the effect of having only many weak relationships, and will be either:

- 1, if at least one contributor has at least one strong, active relationship with a senior contact at the firm (i.e. if the highest influence score is above 50)

- 0.85, if no contributor has any such influential relationships

1 (if 1+ influential relationships)

OR

0.85 (no influential relationships)

Example
  1. A group of data contributors collectively knows 153 people at a consulting firm. Nearly all the relationships are nominal, at only a few points apiece. Sam and Thomas are the only contacts any data contributor has a real relationship with, and neither Sam nor Thomas is senior. The group's overall score with the consulting firm is a 73.
  2. The contributor who knows Sam best is Nadine, who knows him at 82. They continue to work together, increasing their score to 100 (and the influence to, but not above, 50). The group's score with the consulting firm increases to 79.
  3. Sam involves a handful of his colleagues in his interactions with Nadine - including his COO. Nadine now knows the COO at a 10 and her scores with the other few contacts increase similarly. However, there are still no relationships that are both strong and senior, so the group's score with the consulting firm is capped at 85.
  4. The contributor who knows Thomas best is Uri; their relationship is at a 72. Thomas gets promoted to head his department, increasing the influence of the relationship from 36 to 52. Because Thomas was promoted and the highest influence score is now greater than 50, the group's relationship with the consulting company is now 100.

Note that in the above example, even though Nadine has some relationship with the COO and knows Sam better than Uri knows Thomas, Uri's relationship with Thomas is the one that counts for the most influence.


How Did We Do?