Routes & Network: In conversation with IndiGo’s head of Network Planning

In this freewheeling interview with Abhijit Dasgupta – EVP Network Planning and Revenue Management at IndiGo, we go into the depths of IndiGo’s network planning

I would like to thank you, Abhijit, for finding time to chat with me so that we can have our readers understand more about airline Network Planning. In all my trainings so far and even people in the industry belonging to other departments have always been eager to understand network planning.

Ameya: I would really like to understand your career path. How did you end up in Network Planning? Was it by choice or chance and how has your career journey been so far?

AD: Network Planning wasn’t a planned move. I have been in this business for 22 years now, and I joined the aviation industry (Jet Airways) straight out of Business School. After the usual management trainee rotations, I was fortunate enough to get placed in the CEO’s office, and one of the first projects I was assigned to work on was related to engagements with a government body called the Naresh Chandra Committee way back in 2004.

A large part of that engagement was related to allowing private carriers to fly international at that time. Eventually we were allowed to fly international and it kind of snowballed into a role of project managing the company’s international expansion. Jet was transforming at that time, moving from a small domestic carrier to a larger carrier and suddenly larger in network scope. This exposed me to operating dynamics, fundamentals of route modelling, economics, fleet economics, and regulatory dynamics in terms of traffic rights and licensing, which was education on the job. This role was in place for the first 20 destinations for Jet, which included route startups, fleet induction, wide bodies, and so on. This naturally progressed into a full-time international network planning role, and at that time I was the first member of that group. Over a period as I grew, I absorbed leadership of domestic planning, followed by scheduling, partnerships, and strategic regulatory engagement teams, and that was the portfolio I was managing till the end.

It’s also the portfolio that I was fortunate enough to join IndiGo with, and over the last two years I have been privileged to also be given the responsibility to lead the revenue management group.

Ameya: That is very interesting indeed. Coming straight to Network Planning. Seven international destinations announced. How do you typically look at the route, a new destination? Can you explain what a typical process that one goes through?

AD: New ideas could be added, modified, or deleted. These could be for new routes, new stations, or new frequencies. We also look at things that are not working and need adjustment or rationalisation.

These ideas come through to us from several different sources. A lot of it, of course, comes from our own monitoring process. How are point-to-point markets performing? What kind of traffic flow is getting generated in the network, and does that give us clues on future markets? We tap into some industry databases. We look at data and trends available on traffic using other modes of transport. We also get requests and feedback from the environment around us. State governments, industry bodies, and so on. A lot of international tourism authorities come and talk to us; sales teams give us feedback. A new source of learning was actually charters during COVID. It was a difficult time for the industry, but a wonderful time to experiment and discover new markets. This set of ideas forms a hypothesis, which we will then put through a regimen of analytical rigour and economic testing before finalising what to do next on a route.

Ameya: Does the process for International and Domestic differ?

AD: Fundamentally, it remains the same—a typical estimation process—some stimulation is assigned to markets that are either not served before or underserved, an approximation based on learnings from similarly profiled markets that already exist in our network. Short- and long-term economic performance estimates are created based on estimated yields, along with some sort of risk profile. Many new markets are undiscovered due to a lack of visible data, so network planning will always have an element of experimentation. International routes tend to have a much higher connecting traffic composition, so the sets of markets that you need to look at are more diverse when you’re analysing operations on a specific route. But while the models, stimulation factors, growth rates, traffic flows, etc., we would use would vary from market to market, the process of economic modelling will remain the same regardless of market type.

Ameya: How many stages does the process go through?

AD: Many! We will put it through some economic and mathematical rigour to filter from ideas to good ideas (the ones we want to execute), and then we will push forward (if we decide to) to the schedule design process. Here it will get integrated into preexisting schedule design, and that schedule design has its own metrics—fleet efficiency, crew efficiency, airport feasibility, and more—before it moves into the implementation phase where (in case of additions and modifications) we apply for slots, parking bays, traffic rights, schedule and route approvals… and whatever else the case may require and then take the route live. Once any new market, route, or frequency goes live, it will be monitored for operational and economic performance… and this will then create a feedback loop that will drive the entire process over again.

Ameya: Thank you for explaining these in detail. You have a fairly large order book unlike when I did network planning until a decade ago. With deliveries known for the next year, when do you typically start planning for the winter schedule of 2025?

AD: Typically we have two or three different planning windows. We have a long-term planning window where you have a five-year or a seven-year plan that rolls forward every year, which includes a high-level capacity & frequency plan for a set of potential markets. Then you filter that backwards into a midterm (let’s say) three-year network plan, which has more details in terms of schedule design, implementation timeframes, alternatives, etc. We then will look at more immediate issues such as slot availability, traffic rights, and all such things, which drive close-in decisions on what to prioritise or deprioritise.

So, to answer your question, we have a long-term set of objectives and a more detailed sketch of what we want to do in the next three seasons. Then when we start getting closer to the date, we start refining our schedule design, refining our fleet deployment as exact fleet count becomes clearer closer to the date, exact slot availability becomes clear, and so on.

Ameya: Like we have a sales funnel, where out of 100 prospects, 10 get converted to customers. How is it for routes? How many routes are evaluated and thrown aside before one is finalised?

AD: The answer to that differs a lot from case to case. So in the case of Mauritius, we first evaluated Mauritius as a market, then looked at what kind of city pairs we want to operate, filtered through four or five options, and then, of course, we know the range of our plane. The second filtration was which one serves the best based on O&D (Origin-Destination) demand. Which ‘origin’ makes more sense to serve as a pipeline to aggregate connecting traffic flows? This led us to identify Bengaluru as our first destination. Sometimes, it gets narrowed by where the aircraft is available. As I mentioned, each individual case will have different such considerations, which will all get factored in. But we will start with several hypotheses on what needs to be looked at and will filter down to those that represent things worth moving forward with.

Ameya: How about the evaluation of frequency increase or decrease?

AD: It is a very common problem we deal with. Modifications to our existing network, the routes that are not working or need either repositioning in terms of schedules or an increase in frequency. The filtration of which route comes first depends a lot on what time of day the aeroplane is available, for how much time, what the available alternatives are, and then more schedule design specifics.

By and large, we have been very fortunate. It has been a pleasant surprise how many routes we have experimented with that have actually been successful. Definitely, there are things that we have tried and not worked alright, but we are quite surprised with how many things have worked relative to the risk we estimated. It speaks to the significant untapped potential of this market. Of course, in our case we have the benefit of a strong balance sheet, which gives slightly higher risk appetite to try new things. This lets us bring a lot of growth by creating new markets rather than any other strategy. But at the same time, taking such risks requires the rigour of continuous monitoring and the discipline of taking corrective decisions when required.

Ameya: Continuing on that note. Would you want to quote any route which completely went beyond expectations? You can skip the other side of the question on which did not, those are one of the closed routes.

AD: (..laughs..) We had a lot of hope for the CIS markets. We started with quite a few CIS markets last year, but we did not expect them to pick up as fast as they did. We started with less than daily frequencies in all four markets, and the response was very, very strong. It was far stronger than we anticipated. So we had to actually move deck chairs around in forward fleet allocations to take them daily earlier than we had planned because of how strong the response was.

Ameya: I have been a big proponent of Indian carriers flying to CIS since 2017! This gets me to another question which was further down the line. As a former network planner myself, I have been wondering what is with this 0200am arrival in Vietnam and CIS. We have always seen mid-day arrivals like Bali or Mauritius. These markets are new, and apart from connecting it to some 20 points in India, how does this work?

AD: Well, Connections is a big part of that reason. As wonderful as it would be to arrive around hotel check-in, it is unfortunately not how it plays out. It is point to point versus connecting compromise. If we service specific preferred schedules, you don’t connect well to anything in India, and you rely entirely on demand dynamics of the local point-to-point market. This exposes you to the seasonality of the local market, which is of unknown size. So it was a compromise we had to make to ensure that we connect as many markets as possible, thus creating a stronger case for demand meeting the seat capacity we would be placing in the marketplace. Secondly, for us, everything must meet essential metrics of efficiency, because we believe that keeping our unit costs low is important. Our relatively healthy balance sheet gives us the risk appetite to try new things. But to keep the balance sheet healthy, we need to deliver profitability. And a key factor to profitability is unit cost efficiency. By flying in the night, it helps maximise utilisation, so overall we are quite happy that it is working out.

Ameya: Right and now let me shift to something else. If somebody has to be a network planner today, what should a person know or how do you evaluate somebody who wants to be a network planner?

AD: (..laughs..) I am the worst person to ask this question because I am probably the only one who has been in network planning throughout my aviation career. Generally, I think you know the first thing that you need is a very strong foundation in how an airline works. How several of the components of airlines work. You may be an expert in one of those areas. You can take several pathways, say, via revenue management, sales, operations control, or airport management. But I think overall you need to have at least a fundamental understanding of how these different things come together.

As network planners, and you will agree with me, the biggest responsibility you carry when it comes to your own organisation is that you are the originator of the plan that everyone else designs their plans around. What you do affects the plans & effectiveness of a lot of areas of the organization. Planning is one of the few groups that can significantly influence levers of both cost and revenue. So you need to be able to think multidimensionally and need to have an extreme interest in this business, understanding markets, understanding competition, and operational dynamics, and have a broad and open-minded view.

Next, you need to have an interest in, and training for, analysis and data-based decision-making.

Lastly, while you shouldn’t be afraid to experiment, you need to be ruthlessly practical and unemotional in how you approach and measure the success of such experiments.

Ameya: Absolutely! Unemotional as you said. No flights to my hometown unless they make money. So how large is the team today for a 400 aircraft airline? What are the typical sub-teams within Network Planning?

AD: The overall team for Indigo is actually quite small. We’re about 35 people, but that includes the entire planning life cycle. So it includes planning, analytics, route performance, schedule design, and schedule management.

Ameya:What are the typical sub-teams within Network Planning?

AD: Three or four subgroups. The first one is what you would call Network Planning & Analysis.

The team who will evaluate routes, evaluate lead performance, and meet overall network performance. These are owners of the long-term capacity plans. Another team is Schedule Management which will run the operating schedule, publish, distribute and engage with partners for schedule synchronisation, including Slot Management and Parking management. Schedule Design is actually a shared process between Planning and Schedule Management. So Planning will carry the design process for the first leg of its life, and then scheduling will take over somewhere in the middle as we get closer to schedule slot filing deadlines, implementation, and go-live. So optimisation is a shared responsibility.

The third group is Partnerships, who own our relationships with our airline partners and manage commercial engagements, technical integration, and performance monitoring of these relationships.

All these groups work closely with regulatory strategy, who will not only engage with government on traffic rights but also on aviation policy, foreign licensing, and stuff like that. We also work very closely with an airport strategy group that engages with key airport partners.

Ameya: Very interesting, and how many requests do you get for lateral shift to network planning on an average?

AD: (..laughs..) Well, yeah, there are quite a few, of course, and it is an area of interest. As I said, team members come from a lot of different teams within IndiGo as well as from outside of IndiGo.

Ameya: For Network Planning, data is just part of the story. Understanding it, making financial and operational sense is actually the main one, right? Of course you use a lot of softwares to track, plan but there is a lot beyond an excel sheet full of data with a pivot table. I keep saying this for my blog, Knowledge, not information. The information is with everybody, what differentiates is the knowledge. If you could talk a bit on how you find this differentiation amongst people whom you select?

AD: Absolutely. So the first thing we do for folks that join Network Planning is actually overexpose them to data and start asking them questions that force them to start visualising traffic flows. You can relate to this: at the end of the day, flights are simply pipelines that you use to carry traffic flows. We ask them what the fundamental drivers of demand are. People travel for work, to study, to go home, to pray, for holidays, and for medical requirements; there is so much to it.

For a network like ours, we connect people over multiple pathways, so we do a lot of reverse engineering to refine our models. Our size, our scale, and our complexity make us unique, and this will become a big mathematical problem in the future that does not exist in most parts of the world. We are probably going to be one of the very complex airline systems globally.

Ameya: Thank you. That’s a lot of work! IndiGo has this record of not closing a single station. I have tracked it since 2006 and gave up the rotations tracking at 100. Post that something was pulled out, something added, swapped. Now it’s 400+ planes. How much pressure is it to add a station which will not close in future?

AD: Well, frankly we have not run into that problem so far. We did, and I won’t disclose which one, but actually just before COVID we did run into it. We had made a decision, but thankfully the station showed recovery after COVID, and we did not have to take that decision. I cannot rule out that this decision will not come at us at some point in time in the future. There are many markets that we are operating under incentive programs. We know these are sensitive things, but we talk to the right stakeholders and try to take them onboard early so it is not a surprise to anyone. We may actually re-enter markets at a later point when they have matured a bit. It is to be looked at neutrally and practically.

Ameya: Unemotional, as you said!

AD: Haha!

Ameya: When IndiGo launched Varanasi – Bangkok, the internet called it “Paap and Prayaschit” route. The route is not operational any more, but how did the team take it? As a network planner, the last thing on mind is calling a route by a name amidst the melee of analysis.

AD: (..laughs..) This specific example was before my time, but we have had quite a few after, and you laugh at it at the end of the day. Marketing efforts have their own dynamics, and people work hard in coming up with something that will stand out in the crowd, and then the internet can take it to another level, you know!

As planners, you know we look at these things as fun to observe and a light moment in our lives.

Ameya: From the light moments to some difficult ones. You may want to plan 40 flights a day between Mumbai and Delhi but you can’t. The numbers may back them, but the infrastructure or slot constraints won’t. How does it impact your planning process?

AD: Right. A lot goes into engagement and datasets. There are pieces of information that are available and help our decision-making to some extent. Industry data sets that tell times of the day when airports are busy and things like that. Specifically, at very steeply constrained airports, we will not commit to stuff before we have had that engagement and got an understanding of availability.

For a longer-term plan, we can have backup plans, but not for the immediate term. Thankfully, new airports are coming online, so some more alternative solutions should be in place.

Ameya: You made one mention of industry data on popular times. With your experience, what is the most popular time of the day for flights and the most popular day?

AD: Well, look, it’s a well-known fact that no matter which planner or revenue manager in whichever part of the world you talk to, Tuesdays seem to be the weak day or the weakest day. But time of the day is a little harder because I think we end up designing schedules to mitigate a lot of that time of the day variability. Markets have preferences, and so do the types of passengers. A leisure one, a business one, and then the afternoon service carrying a lot more connecting traffic. You are not a good planner or a revenue manager if you are not mitigating that. Red-eye flying, for example, you can intuitively tell is least preferred, but it has its purpose, both economically and from a customer proposition.

Ameya: Airlines from the west announce flights well in advance. KLM, for example, has announced Hyderabad starting Sept 2025. When do you see Indian carriers mature to this level? Even when we are a last minute market.

AD: I think we as an airline and we as an industry are pretty much prepared to go there. We would like to go there. What holds us back right now is that our operating environment is very dynamic. It’s not someone’s fault versus something else; it’s just the way it all needs to come together that leads us to hold decisions for longer than you would have. We are talking to the ecosystem, and this is an evolution that will happen in the few years to come. From an airline perspective and from a customer perspective, of course we want to open flights for sale much earlier. The longer you keep them open for sale, the better the buildup is during launch and thereafter. So it’s in everyone’s interest.

Because India is a late booking market, we don’t end up losing a lot of the booking curve by moving the decision closer in, but that ‘benefit’ has its limits, and in no way does it take away from the fact that all of us in the ecosystem, including our customers, would want to elongate the process. Even from within an airline perspective, our resource planning groups will benefit so much more by having a longer window, which in turn has its own benefits in terms of efficiency and cost-effectiveness.

Ameya: Let’s hope we reach there soon! Of all the questions, the last one is going to put you in soup and I don’t know if you are going to answer it fully. When MOPA was launched, IndiGo shifted a small chunk of flights from Dabolim and added a large chunk. Come 2025, Navi Mumbai and Jewar open up. What many times of MOPA will the response be?

AD: As you predicted, I am going to give you a more generic response! Mopa saw a lot of red-eye flights shifting initially, but we are back where we were at Dabolim in terms of scale, which was constrained, and there was a lot of pent-up demand. So the Goa catchment basically saw a significant growth in capacity deployed as a consequence of the new airport opening.

Fundamentally, that concept will hold for Navi Mumbai and Jewar as well. We are very interested in these airports as fresh builds. I don’t think we look at these as opportunities to rationalise the current primary airports at all. Combined with the local point-to-point opportunity, there is enough connecting opportunity available to tap into both these geographies.

Ameya: Thank you so much, Abhijit, for finding time to sit with me and patiently answer my questions. I am sure the readers are going to enjoy this.

AD: Thank you. It has been good to take a break and talk about network planning, which we generally never talk about!

Special Note

You can support Network Thoughts by ordering Network Thoughts baggage tags !

Follow NetworkThoughts on X (Formerly Twitter), Threads, Bsky, Facebook, Telegram and YouTube.

There is now an option to pay for a coffee to thank Network Thoughts for the effort that is put into the research needed to present these posts. The site will continue to be free. If you have learnt from Network Thoughts and want to pay a fair price for the content, you can scan this QR code to support this website.