Wednesday, April 6, 2022

There was a lot discussed during this session.  The discussion points are grouped below into the large themes of the session.

Assumptions

• How far along Ed-Fi really was
• How much the System Integrator knew and what they were responsible for
• Implementation fidelity evaluations
  • Where are you with data?
  • What are you using data for?
  • How do you access?
  • Once we have the dashboards, needs will be met
  • Timeline assumptions
  • Telephone between stakeholders – not speaking the same language
  • Don’t promise anything to end users on the ground – be honest and candid
  • Wanted to make it more useful
  • What is a teacher’s perception of data
• Quickly identify what to do when assumptions weren’t met
  • Modify skills sets based on assumptions not being met
  • Altering deliverables based on faulty assumptions

Initial Challenges

• Completely overhaul expectations
• Clear that delivering technology on-time is a common problem
• How do you take to the users about using data and making it a daily routine if they really need it – don’t use the product if it’s not useful
• Not set up for success in some ways – had to readjust deliverables

Knowledge deficit

• Cheaper to deploy in university data center – cost comparison
• Hard capping growth
• Knowledge of maintaining infrastructure

Staffing

• Need the people to allow for growth
• 1^st year – trying to process all the problems and triage
• Initial promise to districts kept the work moving forward – didn’t want to let districts down.
• Hire stubborn people who refuse to give up
• We learned what we learned because we were failing, but don’t have two years to see someone to fail so they can learn

Keep going

• Overcome knowledge deficit
  • Package for other districts
  • Go to slack and tracker!
  • Didn’t know how to ask for things that weren’t on their radar
  • Comfortable asking for help
  • Comb through tech docs
• Find the right partners
  • Became fluent in working with vendors and finding the right partners
  • Maintain strong relationships with partners even when it doesn’t go well
  • ClearLaunch – reached out to Arkansas DOE since they used the dashboard too -helped clean up the dashboards.

Evaluate architecture

• Relied on initial integrator – modeled it after the MI Data Hub
• Lofty goal that was based on different assumptions
• Take a step back to understand realistic long-term goals
• Could only scale up to a max because of architectural constraints
• Postgres helped scale – needed to shift to onboard more districts

Building Trust

• Being stubborn – holding to deliverables
• Apologizing
• Admitting failures
• Overly sharing failures
• Only one member did not renew
• Low cost for collaborative members $1/student per year with onboarding fees
• Points in the future where need to evaluate pricing
• Uncertainty during the pandemic – developed new visualization with more flexibility
• Early adopters keep the low cost – wasn’t hurting districts to keep them engaged. They could wait and see
• What you are selling is not the tech – you’re selling you (Rosh and Molly’s experience). Parse and package the experience. You are establishing the hub for all the Ed-Fi knowledge – you aren’t just any other vendor (differentiator). You are the go-to for all things Ed-Fi

Speed Bumps

• Realized the economies of scale by meeting with vendors and not one on one meetings for each district.

Pivot

• Return value
  • Expertise of vendor engagement and management
  • Help SIS vendors so they see the value and can expand their offering and make better use of their products
  • Districts have expanded their SIS use cases that can broaden Ed-Fi use (discipline use case where every district was maximizing use of discipline data)
• Break prior assumptions
  • Working with the state to ensure alignment
• Evaluate change vs. staying put
  • IF the dashboards aren’t working, pivot to building own that are simpler and more sustainable.
  • Package that is more maintainable
  • January 2020 – reviewing scope of work to do customizations with system integrator ($90,000) not going to work with $1/student model. Not financially possible.
  • SQL statements to districts for them to see what they want to see. Could see what the technical folks could do and what the districts needed.
  • Point was to easily add an assessment that could be built and maintained in-house
  • Don’t be afraid to cut your losses
• Maximize what you have
  • Difficult to acquire things with Indiana Univ – procurement policies
  • Where can we spend money that would be sustainable
  • AG grid and fusion charts – cost effective solutions – long term investments
  • Software lives on top of the ODS to allow use of baseline use cases – discipline, attendance, early warning (traffic light model)
  • Continue to hit the ODS limitations – but wanted to keep it closely tied to ODS for near-real time. Grown in this area.
  • Transferred enough knowledge to the districts so they understand when pivots need to happen. Building trust helps with this too.
  • Loaded data in .2 seconds using .NET to cache versus 500 API error with same loads
  • In the meanwhile while we look at future state, districts get the value with what they have been given
  • Rapid development where you can maximize dollars
  • Acknowledge when you need to refocus
  • Minimum proof of value to see the potential and create more offerings (PowerBI visualizations with EdWire)
  • Unexpectedly stumbled upon better technology by staying in the game and take advantage of advancements in technology
  • Internal capacity building allowed for more efficient use of support hours

Flourish

• Build bigger collaborations
  • Understand that districts have different needs and skillsets
  • EdWire, NearShore Devs, etc.
  • Able to learn and transition to Suite 3 while still delivering value to districts
• Vendor trust
  • 8 SIS vendors that trust INsite to have honest conversations
  • Work with your vendor to problem solve together
  • Serve as mediator between vendors and districts
• Stakeholder faith
  • Not just vendors – all stakeholders
  • Triangulation of conversations
  • Can’t just do your own business in a silo if you are going to be the hub
• New opportunities
  • Districts come to INsite because INsite has problem solvers
  • Keep ahead of what the state is doing to plan for district’s roadmap
  • NM PED to build out statewide implementation – guide and scale and vendor management
• Push Community
  • Be vocal about failures – how discipline data was showing up (no description) – certification didn’t mean what they thought it did
  • How do you have the conversation to vendors when they aren’t certified, and you need the data? Should we change what certification means? Push to reform certification.
  • Multi-tenant allows to see the differences between vendors to alleviate burden on vendors – hunting the nulls.
  • Pushing the SIS boundaries of what the model allows for

audience Questions

• The null issue is likely due to SEA implementations versus LEA’s use
• Descriptions are not a tech issue – lack of understanding of how LEAs do things. Surfacing the problem to the vendor.
• More likely to get what you want if you create what you need
• Pricing model - $1/student doesn’t seem sustainable – how? Indiana University costs taken care of and MSDF grant helped kick start. Takes time for districts to understand value. Probably at some point will have to increase for new districts. Diversifying income – be innovative with opportunities (NM DOE). Creative with expenses - Onboarding and technical support – streamlined because of knowledge gains.