How to Optimize Revenues Using Dynamic Pricing?

In August 2023, the worth of onions was Rs120 per kg. What led to it? Crunch in present ensuing from exterior environmental elements and a mild demand. The market, the shopper, the seller, demand, and supply determined the worth. The equivalent goes for a lot of merchandise we buy and promote instantly: movie tickets, bus tickets, e-commerce, fuel, and so forth.

Throughout the concept of worth, demand and supply dictate the prices at which gadgets and suppliers will commerce. When customers’ funds for gadgets and suppliers align with the marginal worth of producing, we receive the optimum market worth, moreover referred to as the aim of equilibrium between demand and supply. Setting the suitable worth on the right time is quintessential for enterprise progress. Pricing managers thus focus on getting close to the “Right Price,” which could also be achieved by way of information and analytics.

Components Influencing Pricing

• Organizational elements: Product stock obtainable, funds constraints.
• Promoting mix: Stage of its product life cycle, Product, Worth, Place, and Promotion.
• Product worth: Value of producing and raw provides.
• Demand for the product: Demand for the companies or merchandise.
• Rivals on the market: Competitor pricing to an enormous extent, determines inside pricing.

What’s Dynamic Pricing?

Dynamic pricing makes use of present tendencies, real-time purchaser conduct, present and demand, and opponents pricing to judge the worth of merchandise provided. It permits gadgets to be provided at fully completely different worth elements, such that purchaser satisfaction is met and corporations can thrive.

Dynamic pricing is adopted when demand is elastic. Dynamic pricing can’t be adopted when demand is inelastic or fully inelastic. When prospects are extraordinarily delicate to changes in worth, there’s a extreme Price Elasticity of Demand, and this property is exploited by way of DP.

As an example – In Bangalore, when only one autorickshaw is obtainable at a selected time and express location on a moist day, a purchaser ready to pay larger (twice or thrice the going value – elastic worth) will get it, whereas one different purchaser who isn’t capable of budge should take the BMTC bus the place prices keep fastened(inelastic).

What are the Targets of Dynamic Pricing?

• Elevated revenue, revenue, flexibility, market share, and purchaser satisfaction.
• Reduce earlier inventory, primary to raised utilization of warehouse home and property.
• Balancing present and demand.

Due to this fact, the success of dynamic pricing is the maximization of Earnings/Earnings/Functionality/Market share/Purchaser satisfaction. Occasion – If in 2021, with out dynamic pricing, 1M fashions had been provided, and the pure trajectory for 2022 is 1.5M fashions. Deploying dynamic pricing, fashions provided ought to enhance to 2M with out shedding out on purchaser NPS or completely different pricing indexes.

Merely put, the YOY enhance in Earnings and Objects is the final phrase success metric of any dynamic pricing algorithm. 

For an AB experiment on dynamic pricing, the success/output metrics that could be thought-about are:

• Widespread order value (AOV)
• Conversion value (CR)
• Earnings per buyer (RPV)
• Gross margin share (GMP)

Components Influencing Dynamic Pricing

• Present: When present is lower, prices are larger.
• Demand: When demand is larger, prices are larger.
• Inventory ranges: Prices are dropped if a listing is earlier and out of vogue. Occasion – clearance sale.
• Purchaser alternative: Ola mini, prime, prime plus, and so forth., have fully completely different pricing tiers.
• Seasonality and festivals: Airtickets all through festive skyrocket, and corporations income from extreme purchaser demand.
• Location: Touristy areas have larger prices.
• Time of day: Midnight prices are larger than midday prices
• Competitor pricing

Types of Dynamic Pricing

• Segmented Pricing: Pupil low price on Amazon Prime, senior citizen low price on trains.
• Time-based Pricing: Motels and flights in India are larger in October/November(festive season) than in August/September.
• Peak Pricing: Surge worth on Uber/Ola
• Pricing depends on rivals.
• Worth Elasticity: The additional elastic the product, the upper fitted to dynamic pricing. All FMCG merchandise are priced this trend in Dmart/Reliance retailers, and so forth.

Earnings/Yield Administration

One can’t talk about pricing and by no means deal with revenue administration. Optimizing pricing, inventory, and distribution to predict demand to maximize revenue.

• The primary objective of revenue administration is selling the suitable product to the suitable purchaser on the right time for the suitable worth and with a suitable pack.
• Segmentation, forecasting, optimization, and pricing are devices utilized in revenue administration.
• It actually works best when merchandise/suppliers are worth elastic.

Is Dynamic Pricing Approved in India?

The licensed and ethical factors of AI and ML are a lot much less talked about in India, so let’s deal with them.

• Dynamic pricing deceives a purchaser into deciding on a pricing which might not be in his/her best curiosity. Moreover, this may increasingly very properly be discriminatory, so the question is – Is it licensed?
• In India, half 3 of the Rivals Act 2002 prohibits worth fixing.
• The section prohibits any settlement between or “observe carried on, or alternative taken by, any affiliation of enterprises or affiliation of people, along with cartels, engaged in an equivalent or associated commerce of merchandise or provision of suppliers,” which determines the market worth.
• If two occasions collude and set prices very associated or nearly associated prices, then it’s illegal. However when one social gathering does so with out the knowledge of the other, then each social gathering won’t be liable.
• The acceptable path is to have an vital Personal Information Security Act(very like these throughout the EU) that safeguards Indian residents in the direction of predatory firm practices.

Disadvantage Assertion

FlyAirportByAir is a taxi-chopper service in Bangalore that offers taxi service to Bangalore Airport. As a result of the demand is relatively fluid and changes primarily based totally on local weather, weekends, and festivals, they should introduce dynamic pricing to reinforce the best line. Write an optimum pricing function that may maximize revenue given:

• Prebooking begins 100 days sooner than
• The total seats per day is 100
• Demand varies between 100 to 200 per day. Generate demand using a straightforward Python code –>np.random.randint(100, 200)
• To simplify pricing -> Worth = Demand – Tickets provided

Given the Days left to book, full seats obtainable, and demand for the day, uncover the suitable worth for every single day.

## Worldwide Variables
DAYS = 100
SEATS = 100

DEMAND_MIN = 100
DEMAND_MAX = 200

Forecasting demand is the 1st step in fixing dynamic pricing. Demand varies with inside and exterior elements. Time-series forecasting or regression methods will be utilized to predict future demand.

demand_hist = [np.random.randint(DEMAND_MIN, DEMAND_MAX) for i in range(10000)]
plt.hist(demand_hist, bins = 100)
print("suggest", np.suggest(demand_hist) )
print("STD", np.std(demand_hist) 

Demand is predicted using the Random function; the suggest value is 150 every single day seats, and the STD is 28.9.

Occasion

Let’s take into consideration this occasion: D0 is the date of the journey. As people solidify their touring plans close to the date of the journey, demand tends to be larger than the preliminary days(D8).  Though the market demand for D0 is 8, solely 3 seats had been booked; my rivals take within the remaining.

Offered that demand is linear, the Python illustration of the equivalent:

def linear_demand(days_left, ticket_left, demand_level):

  tickets_sold_per_day = int(ticket_left/days_left)  
  worth = demand_level - tickets_sold_per_day   ## ticket_left/days_left nearly is 1. 
  return max(0,worth)#import csv

Carry out to calculate revenue:

def cumu_rev(days_left,
             ticket_left,
             pricing_function,
             rev_to_date = 0,
             demand_min = DEMAND_MIN,
             demand_max = DEMAND_MAX):

  if days_left > 0 and ticket_left >0 :

    demand = np.random.randint(demand_min, demand_max+1)
    p = pricing_function(days_left, ticket_left,demand )
    q = demand - p   # demand is linear Q is tickets provided
    q = max(0,q)
    q = min(ticket_left,q)  ## cannot promote larger than tickets obtainable

    return q*p, p

Given this simple function, let’s calculate the worth and revenue for – Sometime sooner than the journey, and the general tickets left are 3. (On account of demand is randomly chosen, revenue and worth might fluctuate, random.seed(10) could also be outlined to get fastened options frequently)

revenue,p = cumu_rev(1, 3,linear_demand )
print("Full Earnings - ", revenue)
print("Worth Per Seat - ", p)

Given this simple function, let’s calculate the worth and revenue for – Sometime sooner than the journey, and the general number of tickets left is 10. The worth per ticket should be larger because of demand is further( 3 to 10).

revenue,p = cumu_rev(1, 10,linear_demand )
print("Full Earnings - ", revenue)
print("Worth Per Seat - ", p)#import csv

With a simple-linear pricing function, it’s evident that as demand will enhance, worth moreover will enhance. Let’s simulate this and try and optimize the pricing function.

Stimulations Using Pricing Capabilities

Let’s stress test this simple function for 10,000 seat reserving simulations using pricing options 1. linear_demand, 2. linear_adj, and three. linear_opti_variable  and choose the easiest pricing that gives the highest revenue, which is the aim of this prepare

1. linear_demand

• Demand is prediction random.
• Worth is the excellence between demand and tickets provided.
• Due to this fact, if demand is larger, the worth may even be larger.

def linear_demand(days_left, ticket_left, demand_level):

  tickets_sold_per_day = int(ticket_left/days_left)
  worth = demand_level - tickets_sold_per_day   ## ticket_left/days_left nearly is 1. 
  return max(0,worth)#import csv

2. linear_adj

• Demand is randomly predicted.
• Worth is linear nevertheless stepwise. An index opti is launched to optimize the sooner linear_demand function proper right into a piecewise function. That when demand is larger, further tickets are booked, which in flip will enhance revenue.
• OPTI is a set value primarily based totally on demand.

def linear_adj(days_left, ticket_left, demand_level):
  """
  As an example we anticipate a number of website guests/views and impressions. 

  If demand is extreme we price larger at fully completely different fees
  """

  if demand_level > 180:
    opti = 3
    worth = demand_level - int( (ticket_left/days_left) + (opti*(demand_level/180)))
  elif demand_level > 150:
    opti = 2
    worth = demand_level - int( (ticket_left/days_left) + (opti*(demand_level/180)))

  elif demand_level > 100:
    opti = 1
    worth = demand_level - int( (ticket_left/days_left) + (opti*(demand_level/180)))

  elif demand_level > 0:
    opti = 0
    worth = demand_level - int( (ticket_left/days_left) + (opti*(demand_level/180)))

  return max(0,worth)#import csv

3. linear_opti_variable

• Very similar to 2, an OPTI index is used, nevertheless this index won’t be fastened, and like Kmeans, the optimum value of OPTI should be chosen primarily based totally on the elbow curve.

def linear_opti_variable(days_left, ticket_left, demand_level, opti = 1):

  worth = demand_level - int( (ticket_left/days_left) + (opti*(demand_level/150)))

  # worth = demand_level - int (ticket_left/days_left)

  ## if opti = 0 then the second time interval turns into 0
  ## As opti elevated second time interval elevated. 
  ## 150 because of on widespread the demand is 150, (100+150)/2
  ## IF demand is larger than 150, then worth will cut back
  ## IF demand is lower than 150 then worth will enhance. 

  return max(0,worth)

Recursive revenue function to calculate cumulative revenue for all 10,000 simulations:

def cumu_rev(days_left,
             ticket_left,
             pricing_function,
             rev_to_date = 0,
             demand_min = DEMAND_MIN,
             demand_max = DEMAND_MAX):

  if days_left > 0 and ticket_left >0 :
    #random.seed(10)
    demand = np.random.randint(demand_min, demand_max+1)
    p = pricing_function(days_left, ticket_left,demand )
    q = demand - p   # demand is linear Q is tickets provided

    q = max(0,q)
    q = min(ticket_left,q)  ## cannot promote larger than tickets obtainable

    return cumu_rev(days_left = days_left-1,
                    ticket_left =ticket_left-q,
                    pricing_function = pricing_function,
                    rev_to_date = rev_to_date+p*q)
  else:
    return rev_to_date

1. Output using linear_demand:

simulation = [cumu_rev(DAYS, SEATS,linear_demand ) for i in range(10000)]
plt.hist(simulation, bins = 100)
print("suggest", np.suggest(simulation) )
print("STD", np.std(simulation) )
plt.title("Earnings For 10K Ticket Reserving")

The widespread revenue primarily based totally on linear_demand function is Rs14,908. That’s evident from the histogram.

2. Output using linear_adj:

simulation = [cumu_rev(DAYS, SEATS,linear_adj ) for i in range(10000)]
plt.hist(simulation, bins = 100)
print("suggest", np.suggest(simulation) )
print("STD", np.std(simulation) )
plt.title("Earnings For 10K Ticket Reserving")

The widespread revenue primarily based totally on linear_adj function is Rs16,146. That’s evident from the histogram.

3. Output using linear_opti_variable:

The 1st step proper right here is to determine on the OTPI value which provides the very best revenue:

opti_mean = []

for j in differ(20):
  simulation = [cumu_rev(DAYS, SEATS,partial(linear_opti_variable, opti= j) ) for i in range(10000)]
  opti_mean.append(np.suggest(simulation))

  plt.plot(opti_mean)
  plt.title("Optimum Value For Earnings Maximization")

def argmax(lst):
  return lst.index(max(lst))
print("The Best OPTI value is -" ,guidelines(differ(20))[argmax(opti_mean)])


>> Output >> The Best OPTI value is - 1

Probably the greatest OPTI value is 1 primarily based totally on the elbow curve. Now let’s uncover revenue for OTPI = 1.

simulation = [cumu_rev(DAYS, SEATS,partial(linear_opti_variable, opti = list(range(20))[argmax(opti_mean)]) ) for i in differ(10000)]
plt.hist(simulation, bins = 100)
print("suggest", np.suggest(simulation) )
print("STD", np.std(simulation) )

The widespread revenue primarily based totally on linear_adj function is Rs15,838. That’s evident from the histogram.

Evaluation of Pricing Capabilities

Based mostly totally on Maximizing Earnings, linear_adj is the easiest pricing function. FlyAirportByAir can test this function and primarily based totally on the AB experiment, its strengths and weaknesses could also be evaluated. Learning from this may be utilized to reinforce effectivity over time.

Conclusion

All through industries like airways, railways, tourism, ticketing, and so forth, DP has been deployed effectively. When carried out rightly, dynamic pricing provides firms with flexibility and a attainable progress lever. With the suitable modifications and elements, DP yields larger purchaser satisfaction. This textual content provides a beginner’s data to the world of DP.

Key Takeaways:

• Dynamic Pricing objectives to optimize revenue, revenue, and purchaser satisfaction.
• The methods utilized in dynamic pricing fluctuate from enterprise to enterprise.
• Probably the greatest methods are chosen primarily based totally on AB outcomes, and iterations and enhancing the algorithm over time.
• It might be utilized solely when demand elasticity exists.

Good luck! Proper right here’s my Linkedin profile if you happen to want to be a part of with me or want to help improve the article. Be comfortable to ping me on Topmate/Mentro; it’s possible you’ll message me collectively along with your query. I’ll be comfy to be linked. Strive my completely different articles on information science and analytics proper right here.

Incessantly Requested Questions

References

• Kaggle Mini Packages: Airline Worth Optimization Microchallenge (https://youtu.be/irjpteecxdg?si=aUH2ifTekQutW-9n)
• Coursera: Fundamentals of revenue administration. (https://coursera.org/be taught/fundamentals-of-revenue-management)
• HBR Analysis: 7 Courses on Dynamic Pricing (Courtesy of Bruce Springsteen) (https://hbr.org/2022/09/7-lessons-on-dynamic-pricing-courtesy-of-bruce-springsteen)
• Dynamic Pricing Model using worth multipliers for on-line bus ticketing platform. (https://www.krjournal.com/index.php/krj/article/view/38/357)
• Dynamic Pricing Strategies for Multiproduct Earnings Administration Points (https://www0.gsb.columbia.edu/school/cmaglaras/papers/multi_rm.pdf)
• Worth Optimisation: From Exploration to Productionising (https://www.youtube.com/watch?v=wPxDibqdg_w)

The media confirmed on this text won’t be owned by Analytics Vidhya and is used on the Creator’s discretion.

Related