Like this post? Share it!
Gregor Herda
Like this post? Share it!

Procuring geospatial services for investment projects: Who, when, and what for?

This is the first part in a series on how best to engage geospatial expertise for  development projects. Much of this is obvious, and partly highly generalised, but maybe there is food for thought here for some of you.

Let’s assume you’re leading, or are a member of, a project team for a Development Finance Institution (DFI). You’re responding to a request for financing. Looking at the general scope, you and your client are keen on applying Earth Observation and GIS somewhere in the project process. There are plenty of avenues you could go down in order to bring the necessary expertise on board. Which road you take will depend, among other things, on:

  1. what kind of project it is, e.g. an investment project, a more upstream Technical Assistance project or a piece of corporate ‘grey’ research,
  2. where in the project life cycle you currently are (true especially for investment projects), and
  3. the level of complexity of the analytical question you’re trying to answer

The early stage: data discovery and exploration

For the sake of simplicity, let’s assume we’re at identification or concept stage of a sovereign-backed investment project.

A circular process diagram consisting of six hexagons detailing the project approval and implementation process in development finance.

At this stage, a common enough task may be to get a first idea of potential sensitive environmental assets or to prioritise sub-project areas based on selection criteria negotiated with the client.

This is an excellent opportunity to draw on individuals with the necessary expertise, either in-house or externally, to dig into the petabytes of publicly available raster, vector and census data. Clients and prospective implementing entities will also possess a wealth of expertise and (possibly not so easy to obtain) geodata to draw on in the process. In fact, they may already have completed a large portion of these analyses as part of their own preparatory work, which you can build on.

This phase is also where the project design can be staked out and improved to a much greater degree than later on in the process, all of which is often negotiated intensively. As such, I firmly believe that this is where interactive data visualisations should be used to a much greater extent than they currently are (which, in my experience, is mostly not at all), as they can really complement your usual fare of static maps and graphs embedded in project documents.

Depending on your willingness to engage with code, these visualisations can include either simple web apps deployed on Hugging Face or JupyterLab Notebooks served to multiple users via JupyterHub. The possibilities nowadays are endless – and either free or at very low cost. This more engaging, collaborative and exploratory way of (spatial) data analysis could lay the groundwork for a better common understanding of fundamental assumptions and objectives throughout the project.

The medium to late stage

Going into preparation, appraisal and implementation is where firms are, more often than not, better placed than individuals to handle highly complex data processing and management tasks. Depending on the use case, either custom or (adapted) turnkey solutions can be appropriate. In either case, the role of well-drafted Terms of Reference (ToRs), be it for ESIA consultants, specialised geospatial service providers or sector-specific consulting firms with some geospatial capacity, cannot be emphasised enough. Good ToRs need to do many things but if I had to pick just three, I would say they need to:

  1. be clear about what analytical question needs to be answered or what use case needs to be addressed,
  2. use the correct technical terminology and specifications, and
  3. be informed by realistic assumptions on budgets and timelines, requiring a deep understanding of the marketplace.

And if you’ll allow me, it may be worthwhile to examine whether the usual Inception Report -> Interim Report -> Final Report schedule of deliverables is really the best fit for these kinds of services. But that’s a topic for another day.

Thoughts? What has your experience been? Please do get in touch and let me know.