How can you measure trust?
Consider a simple equation:
Trusting x Trusted = Trust
In other words: if someone is trusting enough to take a risk (the trustor), and if someone else is trustworthy enough to be worth that risk (the trustee), then when the two parties are a “match” – and you get “trust.”
Suppose you could quantify each. Note that there is more than one way to get the same result for “trust.” For example:
- a “trusting” rating of 8/10 and a “trustworthiness” rating of 3/10 might give a “trust” score of 24 out of 100 – 8×3; and
- a “trusting” rating of 4/10 and a “trustworthiness” rating of 6/10 would give the same “trust” result – 4×5, or 24.
Most of the Data Doesn’t Support Decisions
Most of the trust data out there (think Edelman Trust Barometer, or Pew Research) isn’t about either “trusting” or about “trustworthiness.” It’s simply about the end result, trust. And that’s not very enlightening.
Suppose we get a series of data points about trust, and that they show a decline over time, from 26, to 24, to 20. Does that mean that the trustors got more gun-shy and less willing to trust? Or does it mean that the trustees became more shady, and less trustworthy?
Measuring only the result – trust – is like saying the results of the Yankees vs.Tigers game was 3-2 – without telling you the winner. It’s like saying that the average household income of a small town is $500,000 – without mentioning that one of the residents is a billionaire. It’s like saying that unemployment is down – without mentioning how you count those who are not looking.
If you were to pass laws about regulation – you might want to know the driver of decreased trust. If you were building a marketing campaign – you might want to know which factor shifted. And if you were observing a pattern between two firms, you might want to know why trust declined – was it because of less trusting, or because of less trustworthiness?
There’s a lot more to be said about this rarely observed but simple distinction: let me just point out that there are in fact some sources of data that are actionable and help us get at causal drivers, rather than just identifying results.
The Trust Matrix
The matrix below shows some of these relationships.
1. In the upper left box – Individual Trusting – academics are well aware of the General Social Survey, a fifty-year database with an impeccable pedigree, which permits some fascinating conclusions about our propensity to trust others. Hint: it’s gone down. We’re becoming more and more suspicious in principle.
2. In the box Trustworthy Individuals, the pre-emininent database may be my own company’s Trust Quotient. With over 25,000 data points, a time-proven insight called the Trust Equation, we can now state categorically which gender is more trustworthy, which of the four trust factors are harder drivers of trustworthiness, and the relationship of trustworthiness to industry.
3. What about the critical question of organizational trustworthiness? This is a question we keep trying to answer by reference to trust surveys, which are unable to yield the answer. The best source I know of is Trust Across America’s database of publicly traded US companies (they’re working on expanding it). They have a composite definition well-grounded in commonsense and objective databases, and some compelling data about the correlation between corporate trustworthiness and economic performance.
4. The bottom left box – an organization’s propensity to trust – is something for which I’m not aware of any data. Would someone please correct me if I’m in error? My working hypothesis is that this box has declined considerably.
JP Morgan himself may have lent on the basis of character, but the industry he left behind lends only on secured assets. Except, of course, when they lay off risk through ever-increasingly complex transactions.
Companies routinely won’t even trust small subcontractors, insisting that they self-insure against things like falling on sidewalks. It seems to me that corporations consider a propensity to trust to be roughly tantamount to stupidity. It’s hard to be a trusted organization if you systemically and systematically distrust your stakeholders.
5. Finally, the last column – measurements of trust itself – needs conceptual clarification. When we look at data that says “trust is down,” there are four meanings. We might be referring to trust between individuals, trust between organizations, or trust between organization and individual (with two variations depending on which is trustor and which is trustee).
To Mean What You Say, Say What You Mean
Any of us – not just researchers or academics or survey-takers – can contribute significantly to the discussion of trust simply by being clear about what we mean. If you want to say that bankers have become banksters, then point to data about the decline of trustworthiness on the part of banks – not to composite data that blurs the trustor-trustee distinction.
If you want to say that trust is up in the sharing economy, then use data that talks about the propensity to trust, not just the end result of trustor-trustee interactions.
I have a feeling that some significant chunk of the debate about trust could be improved by simply using clearer language to reflect clearer thinking.
Riding the Shark – Conquering Fear in Selling. New eBook from Charles H. Green, loaded with insights and action steps on how to get back in the selling water, without fear.
Filed Under: Trust Metrics