How I used some simple Technical Analysis to select a Hotel

Whilst down with a severe bout of viral flu, stuck in my study and planning a foreign vacation for later this year, my brain wasn't braining as well as it should have been when it came to comparing and evaluating options. Not only was the location I picked somewhere I had never been before, but it also had literally hundreds of hotels options, many of which were not known chains/brands.

So I turned to my age old friend, Microsoft Excel to put it all down in a sheet, initially just to record an initial few names that seemed the best on my travel site searches as well as being reasonably priced/within my budget and which would help me to go through later when I was feeling better.

Then I had the idea of including Google Maps reviews as well just for comparison. This then spiraled into including the reviews from Agoda , Booking.com and Tripadvisor . Now I had four reviews but some of which were scored out of 5 while others were scored out of 10 and with different review volumes.

So I did what any sane Analyst (but not necessarily normal human) planning a beach side vacation would do and calculated the weighted average rating (doubling the x/5 reviews to x/10 first) based on the score as well as the total reviews.

Based on this my first ranking (pure review based) was

1. Hotel D with 8.7
2. Hotel C with 8.5
3. Hotel A with 8.3
4. Hotel B with 7.7

Now, I had my ranking based on the weighted average reviews for my top 4 picks, but viral flu or not, you can't just make a decision based on reviews, when there's money involved. So the next logical choice was to calculate some kind of Value For Money Index factoring the room rates in relation to the hotels. Because despite all 4 hotels being the same class (as per verifiable online sources) there were still rate differences based on other factors.

For ease, I initially picked just the Agoda rate but needless to say, the pedant in me wasn't happy and so I decided to include the rates from both Booking.com and directly from the relevant hotel website (given how travel websites offer deep discounts etc).

Now to ensure proper comparability, I selected both the most basic room category available (based on the information provided on the hotel website) and without any additional features/frills/addons/pay later/refundable options etc. I.e. purely on a room only basis. I also adjusted for any regional taxes, tourist taxes, levies and charges etc. One hotel had breakfast included in the price and no room only basis, so I had to find the rough breakfast rate from that hotel and remove that from the room rate as well.

After a fair amount of research and adjustments, I was satisfied with the rates I had arrived at from the three different sources and there were no obviously identifiable outliers, meaning after adjustments these three rates were for the most part comparable.

Based on the average room rate, the ranking (in terms of lowest to highest became)

1. Hotel B
2. Hotel A (8% more than B)
3. Hotel D (22% more than B)
4. Hotel C (40% more than B)

But before getting to the VFM score, I realized there was one more cost to factor in, these four hotels were all at different locations across the coast and at differing distances from the airport. Further, the second part of my vacation was in the city and I'd already decided on the hotel there. And so, (as previously mentioned, any sane analyst would do), I decided to factor in a rudimentary travel cost to the rate. As some hotels were much closer to the tourist hotspots on the coast but therefore further away from both the city hotel for the second part of my vacation and the airport.

So I got cumulative distances for

1. Distance from Airport to Hotel
2. Distance from Hotel to hotspots (return trip from the hotel multiplied by expected trips per day)
3. Distance from the Hotel to the city hotel

The next step was to calculate the rough per Km cost from a ride hailing service (for which I downloaded the app) and arrived at the effective cost per Km (not accounting for any kind of surge/peak charges).

Now, I had a room rate, plus an effective total travelling cost as well, with which I was able to calculate my new total room only rate plus travelling cost per hotel across both the same star class and room type.

Based on the effective total cost, the rankings didn't change though the gap between the lowest cost hotel and the others reduced across the board.

1. Hotel B
2. Hotel A (3% less than B, down from 8%)
3. Hotel D (21% less than B, marginally down from 22%)
4. Hotel C (30% less than B, down from 40%)

Now came the final and easiest step and that was to apply the cost to the rating score that I'd calculated above. Which was simple enough, by dividing the cost by the weighted average rating and arriving at a revised score (X), with Lower being Better

Based on this score, the final rankings changed with Hotel A now moving into first place.

1. Hotel A
2. Hotel B (4% higher than A)
3. Hotel D (12% higher than A)
4. Hotel C (23% higher than A)

Now there was one more step I wanted to do and that was factor in some kind of perception index. Meaning, what my perception of the hotel was purely based on pictures, facilities and videos online. Examples included, room design, pool shape (Ugh, this guy is pretentious), beach, garden layout, lounge area etc. Now this was a heavily subjective factor, so whilst this could impact my decision, it should in no way impact the pure technical analysis based ranking we got above.

Instead of just applying a ranking, I looked at a % which simply meant what would my satisfaction percentage be by staying at this hotel, based purely on optics with higher being better. So you don't have to scroll to the top of this already too long article, I'll give the percentages here (Hotel A - 80%, Hotel B - 75%, Hotel C - 90%, Hotel D - 85%). Simply put, based on what I saw and read about the hotels and knowing myself as a consumer, I would most probably get the most amount of utility from Hotel C, then D, etc. After applying that score to the original ranking above, Hotel A still managed to come out on top but with the lead reducing.

1. Hotel A
2. Hotel D (5% higher than A, down from 12%)
3. Hotel C ( 9% higher than A, down from 23%)
4. Hotel B (11% higher than A, up from 4%)

So, based on my perception of the hotels, whilst Hotel A still came out on top, the margins with Hotel C and D changed significantly and Hotel B ended up dropping to last place. As I said, this was a purely subjective measure (I mean I am going to pay for the hotel but still) and so the original ranking of A > B > D > C should be final technical analysis based score.

I have included below a summary how with each additional adjustment how the ranking changed for the four hotels and how Hotel A moved up. There's also a simple average of all 5 ranking methods which doesn't necessarily make a lot of sense given some rankings are linked to others, but I included it anyway

But either way, based on objective and subjective technical analysis, I was able to conclusively prove that without any doubt, Hotel A should be the first and best choice of hotel to book for this holiday in terms of rating, room rate and distance.

And then I booked the hotel my wife told me to book.*

*Which brings me to the most important lesson to learn about Technical Analysis being that no matter how perfect your model is, how rigorously you test your assumptions and how large and/or accurate your sample size is, there can and/or will be a qualitative/undiscounted variable that can change your entire result.