Working at Ushahidi: Pay Equity and Creating a Process of Transparency and Inclusion

(This is part of a series of posts about Ushahidi Operating System (OS) - our management structure)

Today is equal pay day, and in light of this we wanted to talk about the pay equity experiences and processes we have undergone as a company. Two years ago a number of staff bravely brought up the issue of pay equity -- equity across location and experience, but particularly across gender. While we do not publish peoples’ salaries internally, there had been talk and valid concern over whether people of different genders doing the same job were getting paid different amounts. And that was true, there were. The question was, why? Was it due to gender bias, or were there other factors such as location or time working at the company that influenced these inequities? And did we as an organization have a process in place that represented our values?

First we did an internal analysis, to try and determine 1) how were we determining salaries? 2) was that aligned with our values?, and 3) was there in fact inequity in our org? As a remote organization, with staff all over the place who have joined at different times during our nine years of existence, we realized that most salaries were based on a baseline that was negotiated at the time of hire, which was influenced by a number of factors such as: the funds we had at the time, the previous experience or pay someone had at their previous job, and the cost of living in their location. But we hadn’t asked ourselves whether those factors were the most fair and equitable, and our team was right to question that. So as leadership we set about questioning these assumptions.

In our internal analysis we learned a lot about our organization, and I recommend that any organization that hasn’t done this, should, immediately.

We learned that there was a 1:4 ratio between the highest and lowest paid person at the organization; we felt this was a huge accomplishment and a testament to the equity and flatness of our organization. We had a hard time finding comparables, but we found that this is more than twice as good as the norm in the UK non-profit sector (the UK is the only place we could find that published this data), which is 1:8.

The other methodology we used to calculate general compensation equity is ratio of average staff pay to average executive pay. At Ushahidi that ratio is 1:1.56. We looked for a comparison and found that this is considerably more equitable than the norm in the USA for-profit sector, which is 1:243.  Yes that is 1.56 compared to 243, a roughly 155 times larger ratio. But USA CEO-overcompensation is another story for another day. We couldn’t find any data on just the non-profit sector using this methodology, only public technology companies. Microsoft has an impressive 1:11 ratio, and Apple is 1:43 (not taking into account stock options).

We did further analysis on the equity across genders in similar roles at the organization. While doing this, we realized that one of the main factors influencing salaries of people in the same roles was their location - whether it was in a high cost of living or low cost of living location. This insight began to illuminate the complexities in pay equity at a remote organization that has staff in eight different countries around the world. So to get a clear indication of gender pay equity, we had to hold the location and role variables constant. When we did this, we saw that there was equity across the organization, with women getting paid roughly 5% more than men in the same roles and living in places with similar cost of living.

That said, we recognized through this research that while the outcomes we were getting demonstrated pay equity, the process itself was ripe for unconscious bias and its lack of transparency had created a culture where people perceived inequity. And the takeaway here should be that how people feel is all that matters. Regardless of the fact of whether there was or was not pay equity at Ushahidi, the process and lack of transparency of how salaries were determined was making people feel and believe there were vast inequities, and that is toxic.

Research is great, but we now knew we needed to change the way salaries were being determined and become more transparent and less susceptible to any potential unconscious bias.

Here we need to give credit where credit is due. We want to give a big shout out to Buffer for their work on this. Their transparency and writings on this subject influenced us a ton, particularly their openness about their pay calculator and how it was created. A pay calculator is a tool that calculates pay based on structured inputs, leaving bias and negotiation out of it. The question then was what inputs did we value and how did we want to weight them? Luckily, Buffer had published their calculator, so we were able to use that as a starting point. We put forth Buffer’s pay calculator to our team, and together we discussed it. I can’t stress this point enough -- being transparent and inclusive in the process of creating pay equity is as essential as being transparent with the end product.

Together we decided that the key factors that should influence our pay, and seemed most fair were: our role at the organization, our location, our experience in that role, and our loyalty and continued success at the organization

While this was similar to Buffer’s formula, as a team we decided to make some tweaks to their algorithm too: such as removing the additional pay for dependants, and adding in a 10% reduction to the salary data pulled from Glassdoor  as a means of recognizing that as a non-profit covering our costs with a majority of grant funding, we couldn’t afford for-profit salaries. We came up with the following formula, and we ran everyone’s existing salaries through it.

{[(Base Salary *35%* Non-profit Multiplier (0.9)) + (Location Base * 65%)] * (Role Value multiplier 0.9 - 1.1) * (Experience multiplier - 1- 1.3) * (Loyalty/success - 5%/year)} 

Base Salary = Using USA average pulled from Glassdoor

Non-profit Multiplier = We multiply the Base Salary amount by 0.9 since all this data pulls from private sector salaries and expectations.

Location Base = pulled from Numbeo at the city level using maximum amounts (aka, rent for downtown 4 person family data + cost of living for 4 person family).

Experience Multiplier = Determined by hiring manager and company leadership. This factor is influenced directly by the role title used. For instance, if I am the COO, but have never been a COO before, my experience is Beginner (1).

Role Value = this adjusts based on value. Based on Glassdoor the USA the average "Community Manager" is paid $50,000, while a "Customer Service Director" is paid $110,000, so we use this to adjust accordingly to the value of the role in our org.

Loyalty/Success = an annual 5% increase if an employee meets their self-determined Objectives and Key Results.

What we found was pretty interesting. First, that 80% of the team were within $1000 of the calculator. This was great, as we felt that it meant that our calculator was close to accurate. We made adjustments upwards for those who were underpaid, but we didn’t adjust anyone down, but instead let them know what the calculator was showing and that their loyalty increase would go towards bridging the delta until they surpassed the output.

To close this up I would like to leave you with three takeaways we can share from our experience:

A feeling of pay inequality can be as toxic as actual pay inequality, and is the fault of leadership if anyone on their team feels this way.

Creating a culture of equality requires transparency and inclusion throughout the whole process, not just in the final outcome. Do not cloister yourselves with consultants and then pop out with a plan, engage the whole team in the process from start to finish.

Use a transparent algorithm that has the buy-in of the whole team. That algorithm can include both objective inputs like Glassdoor salary averages and Numbeo cost of living averages and also allow for management discernment in experience and loyalty/success metrics, just be transparent about it.

Personally, one of the most meaningful moments of my time as a manager and leader was recently while looking to hire for a software developer and one of the staff members who had originally been brave enough to bring up her concerns a few years back asked for the link to our internal pay equity analysis and calculator because she wanted to, “add the fact that we did that [the pay equity analysis and calculator] as one of the reasons I love working here as I share the job to my female networks.”