In the first week of the project, our team interviewed many of DESCOSUR’s (DESCO South office) staff in order get a good picture of the current state of play. This was covered in depth in our initial blog post a couple weeks ago.
In the second week we focused more on DESCOSUR’s requirements, but additionally had an opportunity to meet with the developer of their current accounting & payroll system (Marston Software), plus their head accountant from Lima.
The accountants at DESCOSUR are frustrated by intermittent connectivity issues with, and data loss by, their Marston accounting software. Additionally, we noted that the software required the use of thousands of product and project codes. In many cases these codes were required to be repeatedly entered for reasons that were dubious at best. The staff knew many of the codes, but clearly it took quite some time to learn how to use the Marston software efficiently and was resulting in many wasted hours by the team.
The accountants at Lima were further frustrated by the long delay in the full deployment of Marston software. The current iteration of the software had been commissioned 3 years prior. Although the expense reporting and accounting components used by DESCOSUR had been completed, the payroll component used by DESCO head office was well overdue.
We spoke with the programmer regarding the architecture and discovered he has an ad hoc development practice. Although he was very approachable he’d had limited formal development training and was unfamiliar with many development best practices, e.g. automated testing, defect tracking, etc. Quite a bit of work would be required in order to improve the reliability of the architecture, e.g. prevent data loss using an external database and to implement automated testing. Given he was busy supporting and finishing the software, it was very unlikely the software would be extensively updated in the near future.
DESCOSUR / DESCO clearly had an immediate need for a reliable system with components for: expensing, payroll and general accounting. But we sat down with their executives and accounting staff and asked them: what would you like to do that you can’t do now, plus asked them to prioritize the criteria for their current accounting software. From our investigation we found that their top priorities were: data loss prevention, functionality, support and customisability.
From their executives, we found that DESCOSUR want to be able to plan budgets using the software, and to be able to match project financial reporting with project progress/status. The executives currently meet with the project and accounting leads twice a month to review project budgets and status. At these meetings they want to see up-to-date project financial reports and status, and have a dashboard showing project health.
So, it is apparent that they need something bigger than a traditional accounting system. They instead are asking for an ERP system that would provide superior project reporting. Additionally, given their priorities it was clear that the Marston software would need to be replaced by a well-supported product/platform.
We met with the executives again on the Friday of week 2 before lunch to discuss our interim findings, specifically that a new ERP system was required, and get a decision from them regarding whether there were political considerations that would force them to stick with Marsten software or if they could move to a new system. We got an answer and are now moving to provide recommendations to support their decision. (come back and read our final update to learn more)
Finally, during week 3, our team was invited to join Marco Paredes, head of Administration, to visit some field offices so we can get a better sense of both the scope of some of the projects as well as some of the challenges faced by field employees. Our first stop was to an Alpaca and Llama support center where we were given a facility tour (kitchen, training rooms and dormitories) as well as a greenhouse for growing vegetables and corrals for Alpaca and Llama. Training is provided to locals to show them how to better care for animals, use proper tools and shearing techniques, and education on basic metrics such as percentage of baby alpaca fur (more valuable) decline and location as the animal ages. It was clear that given some of the highly illustrative diagrams used on charts that the literacy rate is limited in many remote regions.
We proceeded on to Chivay where we stopped in the field office where the local Descosur team was having a meeting. After brief introductions, we proceeded to their remote farm/training site in Linde where we saw their operations including a nursery used to grow seedlings, green houses, and fields where they grew crops (such as Quinoa). We were invite to participate in their annual celebration and ceremony which coincided with the week of Mardi Gras. For many Westerners, it was quite unusual as the process involved “preparing” an already dead baby animal and “offering” It to the gods along and asking for a good harvest, sufficient rain, and good health for the animals. I’m certain this tradition has lasted for centuries, and the celebration included a meal and lasted for several hours. It was indeed festive with generous use of confetti, talcum powder and silly string!
The next morning was a great treat as we were able to visit the renown Colca Canyon and it’s majestic Andean condors, which have a wing span of up to 10 feet. We later returned to the Chivay office and were able to conduct some basic internet speed tests. One thing that became apparent was the incredibly limited internet speed. There’s not much we can do regarding the internet infrastructure in Peru, but it will be a limiting factor for true cloud based computing to occur in the remote offices. Internet speed shouldn’t be in factor in major cities such as Lima and Arequipa.
Signing off for now,
The DESCOSUR Team (Chris Brown, Craig Chan and Kamal LaBreche)