Interlining
Why to interline
What is there to gain by an IL: if you have signed an IL, both your schedules will "act" like one, allowing pax to transfer between your and his/her aircrafts based on the same set of rules as within your own network (minimum and maximum transfer time, transferability at airport, maximum number of transfers ( = 2 !! ), maximum conidered trip length) and, of course, the connections have to make sense. No one is going to fly from a European partner to the US west coast, then back to the US east coast. Based on my knowledge of pax distribution, the more often your flights appear in the ORS, the more likely it is to get pax. However, if you really benefit from adding another connection that has the 2418th best rating, I don't know.
You really can profit from connections where a high transfer rate is to be expected. For instance, running an airline in Spain usually yields good profits on flights to South America, or a bridge between a hub at the US east coast and a hub at the west coast should work nicely. Not only will the flights between the hubs make a profit, but available connections receive more bookings as well, increasing overall profit.
How many ILs to sign is more of a philosopical question, as - at least to me - the required math skills and tools are not available. Also, since AS provides very limited information on where pax come from or go to, this is more of a guessing game. This might be changed with a future UI, as more detailed transfer statistics might become available. However, the AS team repeatedly stated, that they do not provide all potentially available statistics on purpose.
Basically, you have those signing ILs with literally everyone. Then, you have those signing ILs exclusively and with a certain "distribution".
ILs early in the game
I prefer to keep interlining exclusive and well chosen. Also, I don't sign any ILs early on or only when it's a great offer. This also helps me to determine if my business plan for my airline is stable.
For as long as I have no trouble fillig my aircrafts, there simply is no need for an IL. if you have no capacity left, you won’t gain any transfer pax, if you have little left, IL transfer pax might block seats for your internal connections.
Usually, I wait until I establish medium haul routes before considering ILs. When I start my first medium haul routes, I look for potential IL partners at potential destinations. when I operated out of Istanbul, I, for instance, I looked for a partner based at FRA, CDG, LHR, etc. to offer my pax connections to central and NW Europe. On established servers, you can usually find a partner that's not too big (or IL costs are too high) but operates two or three hubs. you can then send your first medium haul flights fo those hubs and have connections available without having to sign numerous partners.
for your hub in in CCS, obvious choices for coutries to look in for partners would be the US and Brazil.
A crude calculation attempt at ILs
Let's consider the costs: obviously, it is great to have as many connections available as possible, however, you need network planners (automatically hired, as most staff) when entertaining ILs. The number is based on your fleet size (or possibly, number of flights per week, I do not know) and the value of your potential partner. Since both your airlines should grow, so does the number of network planners and thus weekly costs.
No matter how many departures you and your partner have at your respective hubs, the maximum number of potential transfer pax is obviously limited by the capacity of the link between your hubs.
Now, you have costs for the network planners/the IL, that need to be covered by the additional pax = transfer pax, you get from that IL. Those pax will either be fed to the link by you or will transfer onto one of your connections. So now, you can calculate the additional per-pax-cost per transfer pax. You can add that to the per-seat-costs of the feeder flight or connecting flight.
My "rule of thumb"
When evaluating a proposed or planned IL, I use the following rule of thumb:
I look up my average margin, available on the income statement, as it gives me an idea, how much of the ticket price I actually get to keep after deducting all costs. Then I take a look at the "facts and figure" page and check my average revenue per pax number.
I now make the assuption, that every transfer pax off a particular link to/from another player's hub will, at average, pay the average renue per pax for the feeder/connecting flight and my profit is the percentage of my overall profit margin.
For example: my profit margin is 30%, my average revenue per pax is 150$. So I am assuming to make .3 * 150 = 45 $ per transfer pax in profits. Once an IL is proposed, the number of required network planners and the costs for them is being displayed on the interlining page. So, now I can calculate how many transfer pax I need to get off of that IL/link, just to cover my additional costs.
Let's say, the IL costs me 45k / week, then I need 1000 pax / week just to cover pax. You now need to make an assumption based on your own transfer numbers available on the office page as well as experience. If the weekly capacity of the link is 700 ( one daily flight with a 100-seater), you are obviously not going to break even. If you have 4 daily 100-seaters, you have a capacity of 2800. 1000 / 2800 would mean, that 35% of the pax on the linking route need to transfer to your network. Obviously, that depends on number of potential connections (increasing as you both grow), geographic situation and competition, but that is a likely number. I usually aim to have at least 50% transfer pax on those routes, it can be up to 95% in rare cases. However, I only sign exclusive agreements with selected partners, where transfers appear to be very likely.
Risks in signing many IL partners
Now, I see a lot of players that sign with everyone, even, if not their hubs but only two or three flights meet at some remote destination of both of them. And some sign more than one partner per airport without increasing capacity of the link. So now, they have to pay for network planners for both partners, doubling the costs. They usually argue "but I can't fill my linking flight without the second/third/fourth partner". Well, be that as it may, what good does it do to fill the linking flight, if the costs of the additional ILs eat up all the profits and more?
This is obviously a very simplified approach, as it does not reflect the advantages of having so many more connections in the ORS. While the percentage is extremely low for a single connection, the sheer numbers will do the trick. I lack the tools to quantify that. Also, the flights operated on a particular link will generate profits all by itself. This needs to be figures in when estimating the total profit potential of a flight as well as interlining.