Audit Ready Data
There is no such thing as Audit Ready Data. To be more precise, there is no such thing as Non-Audit Ready Data. All data is ready to be audited. This blog therefore isn’t about data really but rather about operational processes. In fact, this blog isn’t about auditing either but rather about consistency and coherence. It’s about the consultant’s laptop in the Earth Observation sector and why that is a problem.
The consultant’s laptop is bottled water
You have a problem, you call a consultant. They come in with their laptop, have a chat, make notes on their laptop. Two weeks later, the consultants are back, with their laptops. They show diagrams, graphs, dashboards, slide decks, the solution to your problem, all built in MS Excel. Because that is what consultants do. You are happy, your boss is happy, the client is happy. Everybody is happy. You were thirsty and the consultant brought you a bottle of Apollinaris. You are the man.
Fast forward, 1 year later. Same problem, new data, no budget. The bottle of Apollinaris is empty, sucked dry, in the trash. You will be in the trash soon as well if you don’t find another bottle. Because the auditor is knocking on your door.
But the consultant is gone. She moved from Pink to Blue of the big four. And her laptop is gone as well. With your data, your diagrams, graphs, dashboards and custom XLSX code. All gone.
This is the problem with the consultant’s laptop. And now it is your problem.
This is the problem with a lot of companies now in the EO sector. They are consultants. They satisfy your needs in the short term. There is no structured process behind, data is not consistent over time and space, not even coherent with itself or other data streams from the same organisation.
This is piloting. This is proof of concept. This is not an operational service or Audit Ready Data.
And it’s not just small startups or niche parties. At our cloud provider, they really want us to move to their managed services version of the satellite data crunching engine. They asked us what is holding us back, “just let us know if we need to adapt something in the pre-processing of <sat data source> and we can do that”. MAJOR RED FLAG. Does that mean they are willing to do that for any random customer? At any random, unannounced moment in time?[1]. What they were doing was treating a managed cloud service as a consultant’s laptop.
We still use our own engine.
Turn on the faucet
Many EO downstream service providers nowadays are a version of the consultants laptop. They provide ad-hoc, on demand services, often very local, disguised as SaaS. If you’d estimate how much of the customisation of the end product happens on a laptop, you’d be surprised. The good thing is that the client is happy and gets a positive impression about the service. The bad thing is that it is not traceable, reproducible and thus not auditable.
Many take off with the intention to be a SaaS but not with the mindset. They often get caught in the gravity field of early adopter feature requests.
To move to a SaaS mindset for your product or service you can do the “what if?” game:
- What if the client wants 100 instead of 1 item of this?
- What if the client wants the whole country, the whole continent instead of this pilot area?
- What if I don’t have 1 client but a 100 and each of them calls me on average 10 minutes per week?
In other words: does it scale? Not only the product, but also your organisation. Selling gourmet bottled water is not scalable. Turning on the faucets is.
You don’t want to order or go to the shop for another bottle of water or rely on your storage all the time. You just want to turn on the faucet and have a glass of nice cold water. Everytime, everywhere, always the same. The water might look a little rusty though …
Rusty water
We need 95% accurate data.
No you don’t. What you need is a consistent dataset. A consistent, continuous, reliable stream of data. You want to be able to compare your assets in Cameroon to your assets in Indonesia to your assets in Colombia. You want to be able to compare your 2021 data to 2016 to 2001. You want to see long term trends in your global supply chain. You want to be able to proof that you are doing better than you competitors
Why? Because:
- You have to report on ESG indicators because of EU legislation;
- You want to know if your assets are still there;
- You want to measure the impact of your boardroom decisions.
What you also want is coherent data. You don’t want your global supply chain deforestation risk data contradicting your scope-3 emissions conversion factor estimates.
What businesses want is consistent, coherent data. What businesses want is scalable data that is good enough. Scalability requires consistency and coherence. Make sure your approach is consistent & coherent and scalable, only then aim for accuracy.
If you are looking for a (Remote Sensing) service provider for your reporting needs, look for a provider that delivers:
- Coherent data
- Consistent data [2]
- Audit-ready data
- AND ONLY THEN start worrying about accuracy where good enough is good enough
If 1 and 2 are covered, you almost certainly get 3 for free, and 4 becomes a reasonable requirement. It will greatly help if they are aware of or even have gone through an assurance process with an auditor before. But any organisation with structured processes will pass.
You don’t want to get sick, but accept that the water is still a bit rusty. We’ll get there. Be aware that a reliably running tap is a miracle already.
A thirsty market
When you’re thirsty, it’s tempting to go for a craft beer or a vin nature. But are you out with your mates, do you have a romantic dinner with your boyfriend or are you running a marathon? And what if it’s not just you, but a whole sector, a whole market that is thirsty. Probably you will be looking for something that scales, is affordable and of good enough quality. Like a reliable faucet providing not too rusty water.
Hello SaaS oriented EO service providers. Both of you.
We need EO service providers with a SaaS mindset providing coherent, consistent datastreams of good enough quality (audit ready data) to quench the market’s thirst. It’s ok to patch up evolving market needs with a consultancy layer. But the mindset, the goal, should be audit ready data.
What is important to realise for you, us, me, as a Remote Sensing provider, is that the clients really are not interested in satellites, at all. They don’t even care about your insights or beautiful maps. They want your data to fulfil their reporting needs, to know if their assets are still there or, as a basis for their strategic boardroom decisions. And that data better be auditable.
[1] Yes.
[2] Please note there are 2 kinds of consistency; internal consistency, within a dataset (as used in this piece) and external consistency as compared to an independent standard.