Predict Churn for Cellphone Subscribers

cellphone churn

This model scores whether existing mobile telecommunication (cellphone) customers willchurni.e. whether customers, at the end of thier contract period, will renew thier cellphone subscription or not.

The model construction and its inference can be a precursor to a marketing and sales campaign to entice potential churners to renew thier subscription through sales promotions and other incentives.

cellphone churn

The model is based on a customer's signature consisting of over 15 variables representing cellphone usage details. It is not based on the customer's profile or any other identifying data besides the customer's area code and state.

Management Foresightoffers more sophisticated models that include customer demographics, customer interaction patterns and other variables. See below for model customization options and pricing.

Submitted By
Management Foresight Inc
Cost
$999/month      Click for details  
Date Added
12/15/2013
Source Code
Not Available
Language
C++ and Fortran
Model Type
Classification
Algorithm
Proprietary
 
Data Dictionary
Attribute Name Brief Description Type Range Low Range High Values
STATE State Abbreviations Multi-valued String NA NA AL - WY
ACCT LEN Account Length Number 1 243 NA
AREA CODE 3-digit Phone Area Code Multi-valued Number NA NA 408, 415, 510
PH NUM 7-digit Phone Number Number 327-1058 422-9964 NA
INTL PLAN International Plan Boolean NA NA Yes, No
VMAIL Voice Mail Plan Boolean NA NA Yes, No
NUM VMAILS Number of Voice Mails Number 0 51 NA
TOTAL DAY MINS Total Day Minutes Decimal Number 0 350.8 NA
TOTAL DAY CALLS Total Day Calls Number 0 165 NA
TOTAL DAY CHARGE Total Day Charge Decimal Number 0 59.64 NA
TOTAL EVE MINS Total Evening Minutes Decimal Number 0 363.7 NA
TOTAL EVE CALLS Total Evening Calls Number 0 170 NA
TOTAL EVE CHARGE Total Evening Charge Decimal Number 0 30.91 NA
TOTAL NIGHT MINS Total Night Minutes Decimal Number 23.2 395 NA
TOTAL NIGHT CALLS Total Night Calls Number 33 175 NA
TOTAL NIGHT CHARGE Total Night Charge Decimal Number 1.04 17.77 NA
TOTAL INTL MINS Total International Minutes Decimal Number 0 20 NA
TOTAL INTL CALLS Total International Calls Number 0 20 NA
TOTAL INTL CHARGE Total International Charge Number 0 9 NA
NUM CUST SUP CALLS Number of Customer Service Calls Number NA NA True, False
Sample Data used for train the model
STATE ACCT LEN AREA CODE PH NUM INT PLAN VMAIL NUM VMAILS TOTAL DAY MINS TOTAL DAY CALLS TOTAL DAY CHARGE TOTAL EVE MINS TOTAL EVE CALLS TOTAL EVE CHARGE TOTAL NIGHT MINS TOTAL NIGHT CALLS TOTAL NIGHT CHARGE TOTAL INTL MINS TOTAL INTL CALLS TOTAL INTL CHARGE NUM CUST SUP CALLS CHURN
MS 1 415 408-3977 no no 0 144.8 107 24.62 112.5 66 9.56 218.7 79 9.84 13.8 3 3.73 1 False
TN 1 415 335-5591 no no 0 196.1 107 33.34 296.5 82 25.2 211.5 91 9.52 7 2 1.89 1 False
TX 1 415 396-4254 no no 0 182.1 106 30.96 134.9 106 11.47 152.3 75 6.85 10 3 2.7 5 True
SC 1 408 336-1043 no no 0 123.8 113 21.05 236.2 77 20.08 73.2 81 3.29 3.7 2 1 0 False
AK 1 408 373-1028 no no 0 175.2 74 29.78 151.7 79 12.89 230.5 109 10.37 5.3 3 1.43 1 False
DE 2 415 415-8448 yes no 0 132.1 42 22.46 138.9 88 11.81 192.6 119 8.67 9.1 1 2.46 2 True
TN 3 510 407-8012 yes no 0 161 96 27.37 244.9 82 20.82 180.8 103 8.14 7.7 6 2.08 1 False
TN 3 415 400-4713 no no 0 185 120 31.45 203.7 129 17.31 170.5 89 7.67 14.1 3 3.81 3 False
HI 3 408 355-2872 no no 0 139 99 23.63 250.7 108 21.31 286.2 87 12.88 6.1 3 1.65 4 False
UT 4 510 413-6346 yes no 0 145.3 89 24.7 303.8 93 25.82 206.1 82 9.27 8.9 4 2.4 0 False
AR 5 415 380-2758 no no 0 199.2 106 33.86 187.3 12 15.92 214 85 9.63 13.3 3 3.59 3 False
MO 6 510 350-9994 no no 0 183.6 117 31.21 256.7 72 21.82 178.6 79 8.04 10.2 2 2.75 1 False
OR 6 408 408-1331 no no 0 226.5 93 38.51 152.1 122 12.93 164.4 98 7.4 9.4 4 2.54 3 False
IN 7 415 358-9146 no no 0 206.7 87 35.14 281.1 83 23.89 158.5 77 7.13 11 5 2.97 3 False
MT 10 510 374-5965 no no 0 183 103 31.11 214.8 77 18.26 206.4 73 9.29 8.7 6 2.35 2 False
VA 10 415 352-5697 no no 0 222.2 127 37.77 153.1 125 13.01 227.4 80 10.23 12.9 4 3.48 1 False
VA 10 408 349-4396 no no 0 186.1 112 31.64 190.2 66 16.17 282.8 57 12.73 11.4 6 3.08 2 False
NY 11 415 401-4650 no no 0 143.4 130 24.38 289.4 50 24.6 194 100 8.73 9.7 6 2.62 2 False
ND 12 510 379-5211 yes no 0 216.7 117 36.84 116.5 126 9.9 220 110 9.9 9.8 4 2.65 2 False
OR 12 415 378-4179 no no 0 204.6 98 34.78 212.5 90 18.06 182.1 95 8.19 9.8 7 2.65 2 False
AZ 12 408 360-1596 no no 0 249.6 118 42.43 252.4 119 21.45 280.2 90 12.61 11.8 3 3.19 1 True
WV 13 415 334-6142 no no 0 146.4 74 24.89 148.5 92 12.62 216.7 96 9.75 11.3 3 3.05 1 False
NH 13 415 356-7580 no no 0 193.2 89 32.84 194.4 90 16.52 186.5 104 8.39 9.7 2 2.62 4 False
LA 13 415 388-9653 no no 0 58.4 121 9.93 262.2 64 22.29 159 115 7.15 11.9 5 3.21 1 False
MS 13 415 413-7468 no no 0 303.2 133 51.54 170.5 86 14.49 227.6 80 10.24 11.5 3 3.11 0 True
AL 13 415 354-4333 no no 0 143.1 139 24.33 239.6 88 20.37 221.7 123 9.98 7.1 5 1.92 2 False
MA 13 408 411-4293 no no 0 220.4 100 37.47 211.2 79 17.95 259.3 112 11.67 13.6 8 3.67 2 False
NY 15 510 394-3312 yes no 0 141.4 80 24.04 123.9 76 10.53 323.5 88 14.56 8.1 3 2.19 2 False
MO 15 415 417-9814 no no 0 135.2 101 22.98 152.5 79 12.96 224.8 83 10.12 8.4 5 2.27 2 False
OK 15 415 408-2002 no no 0 121.1 130 20.59 216 86 18.36 235.1 33 10.58 16.1 5 4.35 2 False
VA 16 510 368-2583 no no 0 205.6 69 34.95 169.5 93 14.41 220.1 64 9.9 10.9 3 2.94 0 False
WI 16 510 405-5305 no no 0 229.6 78 39.03 205.7 108 17.48 166.2 91 7.48 10.8 2 2.92 0 True
NM 16 510 367-9259 no no 0 144.8 84 24.62 164.9 141 14.02 231.5 75 10.42 8.2 4 2.21 2 False
MT 16 510 338-1724 no no 0 153.2 65 26.04 229.7 90 19.52 148.2 94 6.67 10.7 8 2.89 1 False
WY 16 415 400-3197 no no 0 174.7 83 29.7 280.8 122 23.87 171.7 80 7.73 10.5 8 2.84 5 False
IL 16 415 342-2013 yes no 0 110 91 18.7 147.3 75 12.52 190.5 73 8.57 6.4 7 1.73 0 False
AL 16 415 336-2322 no no 0 161.9 100 27.52 230.1 138 19.56 148.8 78 6.7 10.2 11 2.75 3 False
AL 16 408 403-9417 no no 0 209.5 89 35.62 172.8 85 14.69 94.1 102 4.23 8.8 4 2.38 1 False
RI 17 415 396-9656 no no 0 161.5 123 27.46 214.2 81 18.21 315 106 14.18 8.6 5 2.32 1 False
MA 17 415 376-4705 yes no 0 162.8 118 27.68 229.6 91 19.52 332.7 94 14.97 13.6 3 3.67 0 True
MD 18 408 347-7898 no no 0 273.6 93 46.51 114.6 116 9.74 250.6 120 11.28 8.2 4 2.21 1 False
GA 18 408 394-6382 no no 0 197 97 33.49 203.7 107 17.31 202 105 9.09 8.7 3 2.35 3 False
SC 19 510 408-5322 no no 0 259.4 116 44.1 269.7 109 22.92 175.3 130 7.89 9.5 3 2.57 1 True
ME 19 415 404-5597 no no 0 201.5 123 34.26 129.2 110 10.98 220.6 98 9.93 12.9 4 3.48 1 False
AL 19 415 380-3910 yes no 0 237.7 98 40.41 207.1 121 17.6 182.2 95 8.2 4.5 4 1.22 0 False
NH 19 408 361-3337 no no 0 186.1 98 31.64 254.3 57 21.62 214 127 9.63 14.6 7 3.94 2 False
MO 20 415 353-2630 no no 0 190 109 32.3 258.2 84 21.95 181.5 102 8.17 6.3 6 1.7 0 False
WI 20 408 344-5967 no no 0 186.8 89 31.76 253.4 51 21.54 273.1 105 12.29 12.3 6 3.32 2 False
NE 21 510 408-3606 no no 0 225 110 38.25 244.2 111 20.76 221.2 93 9.95 10.7 4 2.89 0 False
IL 21 415 343-9658 no no 0 146 78 24.82 109.7 79 9.32 247.4 108 11.13 6.8 7 1.84 0 False
NY 21 415 335-2274 no no 0 244.7 81 41.6 168 117 14.28 281.5 87 12.67 6.6 1 1.78 1 False
WV 21 415 332-5582 no no 0 91.9 109 15.62 198.4 111 16.86 171.7 125 7.73 13 7 3.51 2 False
KY 21 415 412-1991 no no 0 92.6 95 15.74 161.9 70 13.76 285 78 12.83 11.3 3 3.05 5 True
VA 21 415 351-6366 no no 0 223.2 142 37.94 216.5 114 18.4 214.7 111 9.66 12.4 2 3.35 1 False
NV 22 510 393-6475 no no 0 160.4 108 27.27 218.1 88 18.54 192.9 115 8.68 12.5 4 3.38 1 False
FL 22 415 378-9506 no no 0 181.8 108 30.91 198.6 148 16.88 206.6 96 9.3 9.3 3 2.51 2 False
SC 22 408 331-5138 no no 0 110.3 107 18.75 166.5 93 14.15 202.3 96 9.1 9.5 5 2.57 0 False
CT 22 408 345-2401 no no 0 207.7 116 35.31 210.6 99 17.9 238.2 88 10.72 9.6 5 2.59 0 False
CT 23 510 370-5527 no no 0 321.6 107 54.67 251.6 115 21.39 141.1 158 6.35 11.3 3 3.05 2 True
ME 23 510 376-9607 no no 0 113.1 74 19.23 168.8 95 14.35 262.9 126 11.83 6.9 2 1.86 1 True
CO 23 408 393-4027 no no 0 190.2 89 32.33 166.4 108 14.14 219.8 73 9.89 15 4 4.05 6 False
KS 24 510 369-5449 no no 0 243 91 41.31 183.9 77 15.63 184.3 109 8.29 15.3 6 4.13 0 True
HI 24 415 343-2077 no no 0 241.9 104 41.12 145.2 112 12.34 214.5 105 9.65 6.6 5 1.78 1 False
HI 24 415 329-8788 no no 0 235.6 132 40.05 115.9 129 9.85 185.4 136 8.34 16.2 2 4.37 0 False
HI 24 415 398-4431 no no 0 156.2 104 26.55 90 101 7.65 205.1 116 9.23 7.3 5 1.97 1 False
NJ 24 408 393-7826 no no 0 265.6 86 45.15 208.8 102 17.75 182.5 105 8.21 11.1 6 3 2 True
DC 24 408 369-3626 no no 0 149 73 25.33 131 81 11.14 238.6 69 10.74 8.6 3 2.32 2 True
ME 25 510 332-7391 no no 0 242.6 69 41.24 209 117 17.77 219.7 82 9.89 14.4 6 3.89 2 False
NM 25 415 381-2709 no no 0 119.3 87 20.28 211.5 101 17.98 268.9 86 12.1 10.5 4 2.84 3 False
MT 25 415 359-7694 no no 0 134.3 98 22.83 202.3 109 17.2 195.9 100 8.82 12.6 5 3.4 2 False
OR 25 408 422-5874 no no 0 178.8 90 30.4 141.2 72 12 203 99 9.14 8.4 5 2.27 2 False
AL 25 408 337-4600 no no 0 264.9 80 45.03 281.2 66 23.9 166.1 80 7.47 8.4 4 2.27 1 True
NC 26 415 393-3300 no no 0 234.5 109 39.87 216.5 129 18.4 191.6 94 8.62 3.5 6 0.95 3 False
MT 27 510 345-1419 no no 0 193.8 102 32.95 118.9 104 10.11 135.9 124 6.12 9.2 3 2.48 0 False
GA 27 510 403-6850 no no 0 72.7 75 12.36 208.6 117 17.73 65.8 71 2.96 9.9 3 2.67 1 False
NV 27 510 398-7414 no no 0 177.6 121 30.19 296.8 92 25.23 192.9 106 8.68 7.6 3 2.05 3 False
SD 27 510 359-3423 no no 0 226.3 95 38.47 274.3 109 23.32 242.7 119 10.92 8.2 3 2.21 2 True
OR 27 510 355-2840 no no 0 232.1 81 39.46 210.8 101 17.92 165.4 87 7.44 15 6 4.05 5 False
ND 27 415 405-1589 no no 0 227.4 67 38.66 248 115 21.08 61.4 109 2.76 7.8 6 2.11 1 False
NC 27 408 345-6515 no no 0 201.2 128 34.2 227.2 100 19.31 145.8 91 6.56 8.4 3 2.27 2 False
SD 27 408 378-4557 no no 0 236.7 110 40.24 231.9 92 19.71 164.7 85 7.41 12.7 6 3.43 1 False
LA 27 408 348-7556 no no 0 82.6 105 14.04 204 99 17.34 224.2 122 10.09 9.1 4 2.46 1 False
RI 28 510 328-8230 no no 0 180.8 109 30.74 288.8 58 24.55 191.9 91 8.64 14.1 6 3.81 2 False
WY 28 415 392-6856 no no 0 187.8 94 31.93 248.6 86 21.13 208.8 124 9.4 10.6 5 2.86 0 False
MT 28 415 357-9136 no no 0 121.7 48 20.69 125.8 112 10.69 261.6 122 11.77 8.3 2 2.24 6 True
TX 28 415 347-1870 no no 0 159.7 79 27.15 216.7 131 18.42 206.7 116 9.3 9.3 3 2.51 2 False
PA 28 415 334-5223 no no 0 168.2 87 28.59 161.7 92 13.74 192.4 112 8.66 10.1 3 2.73 3 False
ME 28 415 402-5014 no no 0 236.8 102 40.26 167.1 87 14.2 280.2 115 12.61 9.7 3 2.62 3 False
MD 29 510 367-1024 no no 0 195.6 71 33.25 126.4 74 10.74 148.6 87 6.69 14.2 4 3.83 1 False
MS 29 510 401-6982 no no 0 313.2 103 53.24 216.3 151 18.39 218.4 106 9.83 12.8 4 3.46 2 True
WA 29 415 397-7411 no no 0 157.4 122 26.76 145 75 12.33 281.8 92 12.68 9.3 2 2.51 1 False
OH 29 408 402-6666 no no 0 196.8 81 33.46 168 110 14.28 132.6 98 5.97 12.7 7 3.43 2 False
MI 30 510 391-6607 no no 0 227.4 88 38.66 182.5 100 15.51 191.7 134 8.63 12.5 3 3.38 0 False
WV 30 510 411-8043 no no 0 162.3 96 27.59 244 122 20.74 180.1 89 8.1 9.1 4 2.46 2 False
NC 30 510 404-5427 no no 0 145 76 24.65 240.7 112 20.46 197.1 134 8.87 7.1 4 1.92 3 False
NM 30 415 405-8370 no no 0 169.9 144 28.88 225.2 118 19.14 169.7 93 7.64 11.4 7 3.08 1 False
SD 30 415 354-8088 no no 0 247.4 107 42.06 175.9 76 14.95 287.4 90 12.93 11.3 2 3.05 0 False
CT 30 408 410-5192 no no 0 137.6 108 23.39 162 80 13.77 187.7 126 8.45 5.8 10 1.57 3 False
MN 30 408 399-4800 no no 0 54 68 9.18 179.3 96 15.24 247.2 101 11.12 10.2 8 2.75 1 False
FL 31 510 402-3634 no no 0 165.4 84 28.12 203.7 107 17.31 201.7 65 9.08 8.2 1 2.21 1 False
Notes:
  • Only100rows are shown here. Thefull sample dataused to train the model has3333rows.
  • Churn (last column) is the attribute to predict. [True= "Customer churned",False= "Customer did not churn"].

Model Pricing: Details and Options

Base and Custom Models
The base subscription fees lets you use the base model "as is" i.e without any customization.It permits you to score up to 100 datasets per month, with a combined total not to exceed one million (1,000,000) observations. Each line/row in a csv file counts as one observation. The datasets can be submitted interactively using the Interactive Form or by uploading a csv file either through this site or programmatically using the REST APIs.

To model other scenarios a user might want to customize the model, perhaps add a few variables, use different variables, or change the problem definition (e.g. instead of churn modeling, an analyst from the credit card industry may want to model which consumers are likely/not-likely to pay their credit card bills this month, etc). Management Foresight would need to build a custom model for the user in these situations. Customization charges are extra and depend on the type and scope of customization. See table below for options and pricing.
Pricing Tiers
 
Options Base Tier 1 Tier 2 Tier 3
Customization None Add or Change Variables Tier 1 + change problem Definition Same as Tier 2
Hosting On Snap Cloud On Snap Cloud On Snap Cloud (dedicated VM).
or
On Premises.
On Snap Cloud (dedicated VM).
or
On Premises.
Data Extraction and Transformation None
[Clean data expected in exact format]
Same as Base Support for skewed or missing data in on-going data stream.

Support for multiple data feeds/sources.

Support for data preparation to feed into the model on a monthly basis.
On Snap Cloud (dedicated VM)
or
On Premises
Service Level Agreement Standard Snap SLA Standard Snap SLA Standard Snap SLA for Snap Cloud hosting Snap SLA for Snap Cloud hosting
Reports Standard Snap Reports Standard Snap Reports Standard Snap Reports Custom Monthly Reports
Upgrades None None Annual model refresh (with annual contract) Annual model refresh (with annual contract)
Support Standard Snap Support Standard Snap Support Snap Support for Snap Cloud deployment.

Dedicated remote part-time engineer available for a fee
Snap Support for Snap Cloud deployment

Dedicated remote part-time engineer available for a fee
Subscription Fees $999/month Starting at $1999/month Starting at $4999/month Starting at $8999/month
Upfront fees for model development and data preparation None None Starting at $50k
(varies by scope and complexity)
Starting at $100k
(varies by scope and complexity)