ContentSignals

Step 3 of 4 on the thread

Valuation: what a graph asset is worth

Once signals are in a graph, the graph can be valued. The 2012 filing does it with a method marketers have long used to value customer relationships, adapted to graphs. Swomi publishes it as Graph Modeling for Valuation 1.0. Here it is in plain words.

The method in one line

The filing calls its approach "a version of customer equity and customer lifetime value graph assets oriented to social graphs".⁠[1] Customer lifetime value estimates what one relationship contributes over its whole future;⁠[2] customer equity is the total of the lifetime values of all of a company's customers.⁠[3] The filing applies both to the relationships in a graph.

Step one: the lifetime value of one relationship

Take one relationship in the graph: a reader and a creator, or a community and a piece of content. Estimate what it brings in each period from now on. A period further away counts for less, so each period's amount is divided by one plus the discount rate, raised to the number of periods ahead. Add the discounted amounts together and you have the relationship's graph lifetime value, which the filing calls GLV.⁠[1]

CFt
what the relationship brings in period t
d
the discount rate for each period
T
how long the relationship lasts

Step two: the equity of the whole graph

Do the same for every group of relationships, including the groups the graph is expected to gain later, and add the results, discounting each later group in the same way. The filing calls the total graph equity, or GE, and describes it as "the sum of discounted GLVs".⁠[1]

c
a group of relationships: those the graph has now, then those it gains in each later period
Nc
how many relationships are in group c
GLVc
the lifetime value of a relationship in that group

Notation simplified from the filing.

Step three: how likely a relationship is to last

Not every relationship lasts. Alongside its lifetime value formulas, the filing uses a customer retention probability: the chance that a relationship carries on into the next period.⁠[1] A loyal community, whose members keep reading, sharing and coming back, raises the value of a graph. An audience that drifts away after one visit lowers it.

An illustration

Suppose a relationship brings 10 units in each of the next three periods, and the discount rate is 10 per cent. The first period counts for 9.09 units, the second for 8.26 and the third for 7.51, so the lifetime value is 24.87 units rather than 30. The units are abstract: they stand for whatever the graph is measured in, not money, and the figures describe the arithmetic, not anything anyone will receive.

Three periods, discountedThree bars, one per period. Each has an outline 10 units tall. The filled part is 9.09 units in period 1, 8.26 in period 2 and 7.51 in period 3. Together they make 24.87 units, against 30 undiscounted.9.09Period 18.26Period 27.51Period 310 units
Each later period counts for less. The outline shows the 10 units; the filled part shows what each period counts for once it is discounted.

Graph Modeling for Valuation 1.0

Swomi publishes the method under one name and version, Graph Modeling for Valuation 1.0, so that anyone who uses it can say which version they used. Version 1.0 is the method set out on this page, in the three steps above.

A later version changes the method only with a new number, and earlier versions stay published.

swomi.io, the marketplace for digital assets, will use Graph Modeling for Valuation as its reference and will state the version and the inputs behind each valuation it shows. It is not open yet.

What ContentSignals takes from the filing, and what it leaves

The filing goes further than valuation. Its abstract describes social and graph credit scoring, and a clearing-house where graph assets can be held as collateral.⁠[1] ContentSignals takes only the valuation: a consistent way to say what a graph holds, so that value can come back to people as points. It offers no credit, no collateral and no financial service of any kind.

References

Numbered in the order they are first cited on this page; each entry links back to the places that cite it. Every source was opened on 4 November 2026.

  1. ↑ a b c d e US20130290226A1: System and method for social graph and graph assets valuation and monetization. Google Patents. Inventor Maynard Dokken. Priority 5 April 2012; filed 5 April 2013; published 31 October 2013. Status: abandoned.
  2. ↑ Customer lifetime value. Wikipedia.
  3. ↑ Customer equity. Wikipedia.