|
| 1 | +.. -*- rst -*- |
| 2 | +
|
| 3 | +======================================= |
| 4 | +Online Advertising: Campaign Management |
| 5 | +======================================= |
| 6 | + |
| 7 | +.. default-domain:: mongodb |
| 8 | + |
| 9 | +Overview |
| 10 | +-------- |
| 11 | + |
| 12 | +This document outlines the basic patterns and principless for using |
| 13 | +MongoDB as a persistent storage engine for an online advertising network. In |
| 14 | +particular, this document focuses on creating and maintaining an advertising |
| 15 | +campaign with a pre-set daily budget and cost per click (CPC) and cost per |
| 16 | +thousand impressions (CPM) limit. |
| 17 | + |
| 18 | +Problem |
| 19 | +~~~~~~~ |
| 20 | + |
| 21 | +You want to create an advertising network that will serve ads to a variety of |
| 22 | +online media. As part of this ad serving, you want to track which ads are |
| 23 | +available to be served, based on both the daily budget and the CPC and CPM |
| 24 | +limits. |
| 25 | + |
| 26 | +As part of a campaign, a customer creates one or more *zones*, where |
| 27 | +each zone represents some location on a group of pages. Each zone in a campaign |
| 28 | +then has one or more ads assigned to it |
| 29 | + |
| 30 | +Solution |
| 31 | +~~~~~~~~ |
| 32 | + |
| 33 | +In this solution, you will store each campaign's metadata in its own document, |
| 34 | +including budget, limits, targets, and ongoing statistics. This data can be |
| 35 | +modified before the campaign starts or during the campaign itself. |
| 36 | + |
| 37 | +Schema Design |
| 38 | +~~~~~~~~~~~~~ |
| 39 | + |
| 40 | +The schema for campaign management consists of a two collections, one which |
| 41 | +stores campaign metadata, and another for campaign statistics. The campaign |
| 42 | +metadata collection ``campaign.metadata`` documents have the following |
| 43 | +format: |
| 44 | + |
| 45 | +.. code-block:: javascript |
| 46 | +
|
| 47 | + { |
| 48 | + _id: ObjectId(...), |
| 49 | + customer_id: ObjectId(...), |
| 50 | + title: "August Shoes Campaign", |
| 51 | + begin: ISODate("2012-08-01T00:00:00Z"), |
| 52 | + end: ISODate("2012-08-31T00:00:00Z"), |
| 53 | + zones: { |
| 54 | + z1: { |
| 55 | + site: 'cnn.com', |
| 56 | + page: 'stories/shoes/.*', zone: 'banner', |
| 57 | + limit: { type: 'cpm', value: 2000 }, |
| 58 | + ad_ids: [ 'ad1', 'ad2' ] }, |
| 59 | + z2: { |
| 60 | + site: 'cnn.com', |
| 61 | + page: 'stories/shoes/.*', zone: 'tower-1a', |
| 62 | + limit: { type: 'cpc', value: 45 }, |
| 63 | + ad_ids: [ 'ad3' ] } }, |
| 64 | + daily_budget: 25000 |
| 65 | + } |
| 66 | +
|
| 67 | +The statistics are stored in their own collection ``campaign.stats``: |
| 68 | + |
| 69 | +.. code-block:: javascript |
| 70 | +
|
| 71 | + { |
| 72 | + _id: ObjectId(...), // same as campaign ID |
| 73 | + zones: { |
| 74 | + z1: { |
| 75 | + daily_stats: { |
| 76 | + '2012-08-01': { |
| 77 | + total: { impressions: 10146, clicks: 198, conversions: 16 }, |
| 78 | + ad1: { impressions: ... }, |
| 79 | + ad2: { impressions: ... } }, |
| 80 | + '2012-08-02': { |
| 81 | + total: { impressions: 9182, clicks: 183, conversions: 18 }, |
| 82 | + ad1: { impressions: ... }, |
| 83 | + ad2: { impressions: ... } }, |
| 84 | + '2012-08-03': { |
| 85 | + total: { impressions: 9784, clicks: 202, conversions: 21 }, |
| 86 | + ad1: { impressions: ... }, |
| 87 | + ad2: { impressions: ... } }, |
| 88 | + ... |
| 89 | + '2012-08-31': { |
| 90 | + total: { impressions: 0, clicks: 0, conversions: 0 }, |
| 91 | + ad1: { impressions: 0, clicks: 0, conversions: 0 }, |
| 92 | + ad2: { impressions: 0, clicks: 0, conversions: 0 } } } |
| 93 | + }, |
| 94 | + z2: { |
| 95 | + daily_stats: { |
| 96 | + '2012-08-01': { |
| 97 | + total: { impressions: 10457, clicks: 79, conversions: 14 }, |
| 98 | + ad3: { impressions: ... } }, |
| 99 | + '2012-08-02': { |
| 100 | + total: { impressions: 9283, clicks: 53, conversions: 8 }, |
| 101 | + ad3: { impressions: ... } }, |
| 102 | + '2012-08-03': { |
| 103 | + total: { impressions: 9197, clicks: 72, conversions: 14 }, |
| 104 | + ad3: { impressions: ... } }, |
| 105 | + ... |
| 106 | + '2012-08-31': { |
| 107 | + total: { impressions: 0, clicks: 0, conversions: 0 }, |
| 108 | + ad1: { impressions: 0, clicks: 0, conversions: 0 }, |
| 109 | + ad2: { impressions: 0, clicks: 0, conversions: 0 } } } |
| 110 | + }], |
| 111 | + daily_spent: { |
| 112 | + '2012-08-01': 23847, |
| 113 | + '2012-08-02': 20749, |
| 114 | + ... |
| 115 | + '2012-08-12': 0, |
| 116 | + '2012-08-13': 0, |
| 117 | + ... |
| 118 | + '2012-08-31': 0 } |
| 119 | + } |
| 120 | +
|
| 121 | +Operations |
| 122 | +---------- |
| 123 | +
|
| 124 | +TODO: summary of the operations section |
| 125 | +
|
| 126 | +The examples that follow use the Python programming language and the |
| 127 | +:api:`PyMongo <python/current>` :term:`driver` for MongoDB, but you |
| 128 | +can implement this system using any language you choose. |
| 129 | +
|
| 130 | +Operation 1 |
| 131 | +~~~~~~~~~~~ |
| 132 | +
|
| 133 | +TODO: describe what the operation is (optional) |
| 134 | +
|
| 135 | +Query |
| 136 | +````` |
| 137 | +
|
| 138 | +TODO: describe query |
| 139 | +
|
| 140 | +Index Support |
| 141 | +````````````` |
| 142 | +
|
| 143 | +TODO: describe indexes to optimize this query |
| 144 | +
|
| 145 | +Sharding |
| 146 | +-------- |
| 147 | +
|
| 148 | +.. seealso:: ":doc:`/faq/sharding`" and the ":wiki:`Sharding` wiki |
| 149 | + page. |
0 commit comments