Education has always been an important subject for Andrew Ho. In 2009, he was offered a chance to leave the private sector, where he was working as an IT consultant, to join the educational nonprofit Teach For All. Ho welcomed the change.
As the organization's senior director of IT, Ho says he feels privileged knowing his work makes a positive difference in the lives of the students and teachers Teach For All serves.
Founded in 2007 and headquartered in New York, Teach For All employs 102 staff members, who work in the organization's 32 independent partner offices situated around the globe.
The organization's mission is to provide all children with the opportunity to attain an excellent education. The IT and data management teams collect and analyze data to uncover and distribute knowledge and insights that help teachers, staff and alumni achieve this goal.
Data collection is automated and metrics are always up-to-date
Partner organizations can access reports at any time
Robust data visualization tools provide new ways to display information
Teach For All's partner organizations recruit leaders to teach in their nation's neediest areas, invest in training educators on teaching best practices and foster their continued development as they pursue roles inside and outside of education. To be successful in these endeavors, these partners rely on the insights the data team culls from program data to make sound strategy and growth decisions.
Before implementing a comprehensive business intelligence (BI) solution, Teach For All used Excel to collect and analyze over 300 quantitative and qualitative data points. These included:
At the time, this was the best method for the young organization to build a proof-of-concept in order to demonstrate that the data was both collectible and useful. While the process worked, it was overwhelming and incredibly time-consuming.
“We were using tricked-out Excel worksheets,” Ho says. “It did the job, but we knew it wasn't a good solution for the long-term. There were significant and obvious challenges to this method.”
The old process involved sending an Excel file to each partner every six months, where they had to enter data into over 300 different fields. If the staff member in charge of entering the data had numbers readily available, the task could be completed in a couple of hours, but this was rarely the case. Most staff members needed a couple of weeks to finish as they searched for hard-to-find data.
The previous data collection process was lengthy and required many steps.
When Teach For All received the completed Excel files, staff spent at least two weeks validating the data and compiling it into one master spreadsheet. Staff analysts then crunched this data to help uncover insights and trends partners could use to assess their performance. The data was also used to calculate benchmarks for partners to see how they stack up next to their peers in the same regions or of similar size and structure.
The process was slow and tedious, but it did confirm the data collected was useful. Teach For All’s directors decided it was worthwhile to invest in a complete BI solution, and began the search in January 2011.
Ho and his team put together a list of software requirements, which they used to evaluate three BI solutions. “When we first started searching, we weren't looking for a standalone BI solution, but a more robust reporting plugin for the CRM tool we were already using. We had the mindset that we would just use our CRM tool if we could tweak the reporting functionality a bit more,” Ho says.
“It wasn't until we spoke with a CRM consultant that focused on non-profits, who suggested we look at broader cloud-based BI solutions, that we decided to look at comprehensive BI software.”
In addition to being web-based and cloud-hosted, the new BI system had to meet two other conditions:
It took five months for Teach For All to fully evaluate all three BI solutions. In the end, the team decided Birst was the clear front-runner.
“We chose Birst because the solution had all our must-haves and because Birst's entire organization was excited and eager to work with us,” says Ho. “The latter part can't be stressed enough because, as a non-profit with a fairly non-traditional use case for BI, we wanted a company that understood what we were trying to achieve. Birst understood quickly, proactively discounted their services and licenses for us and really went the extra mile in answering our questions.”
It took three months to implement Birst software. Teach For All committed to one week of intensive training, which covered system administration tasks, how to connect data to Birst, data modeling and report and dashboard design. The process was relatively painless, and in September 2011, Teach For All was fully up and running on the new system.
Ho stresses the importance of training in order to take full advantage of all of Birst's features. “There's definitely a learning curve, even for Excel jockeys,” he says. “It's imperative for people to get formal Birst training instead of trying to figure it out as they go along.”
Ho says that he and his team didn't fully understand the full extent of the system's capabilities until they had been given a complete introduction to all it's functions, which greatly impacted the efficiency of the data collection and distribution process.
With the new analytics solution, Teach For All's partner organization data and CRM application data are always synced. This is important for three reasons:
The IT team simplified the data collection process by creating an online form that connected to Birst and that partners can access via a secure login. When logged in, partners can update one, some or all of their data points, any time they want.
Screenshot of web form used to collect partner data
The new data is then immediately uploaded into the CRM and synced with Birst via a nightly batch update, which means reports that are generated always reflect the most current data. “When you think about this and compare it to the original process, the time-saving benefits are huge,” says Ho.
The new data collection process is quick and automated.
In the data team's pursuit to provide the best data, they audited what they were collecting to determine what data was most valuable. While some data seemed useful upfront, the team decided not all data points were necessary, and reduced the number of data points they collect from over 300 to 217 to further improving efficiency. Though Birst analytics didn’t contribute directly to the reduction of data points, Ho says it was the visibility afforded by the technology that enabled his team to make the change.
Previously, Teach For All partners relied on the IT team to create even the most basic data reports and charts, and they had to wait weeks for requests to be processed. Now, partners have access to the data and reports they use most often whenever they want.
By logging in to the partner portal, partners can view a variety of pre-created reports and charts. They also have the opportunity to customize reports by filtering data points by country, year of data, program age, total number of participants, staff size and geographical region.
Filtering lets partners zero in on the data that is most relevant to them. For example, a partner in Argentina searching for teacher recruitment benchmark data may choose to view only data from partner organizations in South America, as that data is more useful to them than benchmark data from East Asian partners. Three of the most-used reports include:
➔ Topline stats. This shows the number of participants, teachers, staff and alumni across the partner networks. Partners use this report to give details to donors and the press. This data is frequently requested, so it's very convenient to have it available at all times.
➔ Recruitment and selection timeline. This report shows the recruitment season and application deadline dates for each partner organization. Not every organization recruits at the same pace or follows the same school year—organizations in the southern hemisphere, for instance, use a schedule opposite of those in the northern hemisphere.
This data helps partners keep track of what other organizations are doing, as well as adjust their own schedules. For example, if an organization received too many applications the previous year, they can decide to shorten the recruitment season by a week to prevent this from happening again.
➔ Funding map breakdown. This is a donation report that shows money raised by the organization as a whole and by partner in a fiscal year, broken down by type (e.g. private, government or corporate donation). Partners use this data to see how they compare to fundraising efforts of partner organizations in neighboring countries.
Screenshot of topline stats report that partners can view any time they log in to the partner portal
Ho lauds the data visualization capabilities of the new BI tool. “Simply put, we can do so much more with this BI reporting tool than we could imagine doing with Excel graphs,” he says.
Charts and graphs can be exported into PDF, PowerPoint and CSV files and manipulated in many ways.
A partner viewing a pie chart, for example, can change the format to a column graph, or use their mouse to outline a section of the graph they want to zoom in on to view more analytical information.
Graph showing total number of staff per country
Zoomed in view of graph showing total number of staff per country
Furthermore, graphs are instantly updated with new data, which the data team and partners find very useful because now they don’t have to update the most used graphs manually.
Ho says that Teach For All underestimated the positive effect that Birst would have on their partner organizations. “By making the data easily accessible to all partners, some staff had a shift in mindset,” he explains. “As a result, a number of our partners have restructured how they collect and store data internally to be more effective and efficient.”
Previously, the data collection process happened only twice a year, so partners had to keep track of these numbers until the Excel spreadsheet arrived. This wasn't easy, as not every partner was well organized. With the new process, however, everything is automated and easily accessible.
Partners make updates as they go along simply by logging into the system, and everyone appreciates the higher accuracy of the data. And because partners now understand that entering accurate, up-to-date data directly contributes to successful decisions and program growth, they're inspired to perform this task as best as they can.
Another unexpected perk is that Teach for All now has new ways to share data. Ho is excited by Birst's rollout of HTML5 support and is currently exploring ways to take advantage of this feature, which will allow his team to embed data visualizations, such as charts and graphs, on internal and external websites.
“We will be able to incorporate data visualizations with other contextual information on our web pages to convey key information to users,” he explains. “Rather than static, these data visualizations will update as our data is updated.”
Teach For All is currently developing a knowledge sharing portal which will bring this content and data together in one spot. A partner searching for information about recruitment, for example, will be able to go to the recruitment page and view documents about how to recruit, an embedded YouTube video featuring a lecture from a fellow partner on recruitment best practices and a data visualization of the previous year's recruitment statistics. All this data will be pulled from different sources, including Birst, to create a dashboard that displays the information in a single location.
In this way, Ho says the new system has helped fulfill his vision for the organization. “I joined Teach For All because I wanted to be part of an organization that was making a difference in society and education,” he says. “My favorite part of the job is being able to see my IT vision for this organization come to life, and the adoption of Birst analytics was a big part of that vision.”